SPEEDRUN APPLICATIONS ARE OPENING SOON! And we’re coming to SF.

Join us at the pop-up speedrun cafe (presented by Mercury) in SF during the first week of the SR007 application window: April 20th–24th! Each guest can snag a drink, a food item, and a limited-edition hat (while supplies last).

RSVP on Partiful using the links below to reserve your spot on one of the five days:

Monday (4/20) | Tuesday | Wednesday | Thursday | Friday (4/24)

This Week’s Big Ideas 💡

💻 Andrew Chen comments on the growing wave of non-technical founders using AI coding tools to build companies: “their iteration speed is 10x.”

🪞 Mirror Mirror AI (SR006) founder Yusan Lin explains why she’s building infrastructure for people to own their own likenesses in the AI era.

🛠️ Bella Nazzari explores job openings for the growing discipline of “gtm engineering.”

💼 Join our talent network for more opportunities. If we see a fit for you, we’ll intro you to relevant founders in the portfolio

Every founder is trying to crack distribution. They’re posting on X, debating whether to start a podcast, and wondering if doing the latest viral trend on TikTok is worth the effort.

Meanwhile, LinkedIn is sitting right there—wildly underutilized and practically begging early-stage founders to use it. Eventually, this channel will be saturated like everything else. But for now, for the willing founder, we believe there are customers to be had.

The LinkedIn meta is driven by two factors:

• The audience is already there, pre-qualified. Every other social media fight is a slugfest between political takes, culture war brainrot, an algorithm that changes weekly, while you’re earnestly posting about your latest feature. In contrast, LinkedIn’s feed is overwhelmingly professional. The people scrolling it are decision-makers, investors, hiring managers, and operators looking for “work” content. If you’re selling to businesses or raising money, your buyers are literally browsing between meetings. manager.

• The algorithm is generous. LinkedIn’s distribution is still in growth mode. A post from a 7,000-follower account can get 850 likes and 184 comments—that’s what happened when Salar Shahini of Sweatpals (SR004) announced his team’s $12M raise. That’s a 12%+ engagement rate! The platform actively rewards original posts, especially from personal accounts (not company pages).

What actually works on LinkedIn

After watching hundreds of founders go through speedrun, here are some tips for making performant posts:

Post your actual numbers.

The highest-performing LinkedIn posts from speedrun founders are almost always the ones sharing real metrics. No one wants to read your thoughtslop pieces about vibe coding. Just follow the example of Daniel Dhawan of Rork (SR005)—he posted that Rork users had created over 500,000 projects with 2,000+ apps published to the App Store, hitting #2 in the US Dev Tools category” and let the numbers do the work. Specificity is the currency of trust on LinkedIn. Vague “excited to announce” posts get polite likes from your mother. Actual numbers get inbound leads.

Build in public, but make it a narrative.

Taylor Offer of Atrios (SR005) makes each post build the narrative around his company. He posts about the emotional arc of founding (”Don’t forget how bad you wanted what you currently have” got 44,000+ likes), the tactical realities of scaling, and meta-observations about the platform itself. His audience feels like they’re following a character in a show. Treat LinkedIn as a serialized story to build the kind of following that actually converts into quality leads. As we’ve discussed before, early sales is about people loving the founders first and the product second.

Use your pre-founder credibility.

Your career before your startup is a feature, not baggage. Eric Berman of Lil Snack (SR002) was a Hulu and Crunchyroll executive before founding an AI gaming company. Paddy Lambros of Dex (SR005) ran talent at Atomico, so when he posts about building an AI recruiter, his network of recruiters and hiring managers actually believes what he says. If you spent years in an industry before founding a company in that space, your LinkedIn network is your first customer list.

Don’t overthink the format.

The posts getting the most engagement from speedrun founders are almost comically simple. A photo with a short story. A screenshot of a metric. Michael Lingelbach of Hedra (SR002) posts about AI video breakthroughs in a few short paragraphs—no carousels, no “10 things I learned” threads, or graphic design. His $32M Series A announcement was straightforward and still generated dozens of comments from exactly the audience Hedra wants. LinkedIn rewards “authenticity” over production value. A real photo from your office beats a Canva graphic every time.

How to survive the trough of cringe

Let’s be very clear: the first few LinkedIn posts feel terrible. You’ll write something, hover over the post button, and wonder if your college roommate or former boss is going to think you’ve lost your mind. Call it the “trough of cringe.” Every founder who’s now crushing it on LinkedIn went through it.

The discomfort is a competitive advantage. It’s the barrier that keeps your competitors from doing this. If posting on LinkedIn felt natural and easy, everyone would do it and it wouldn’t work. So start with what you know is true. Write about something you genuinely experienced this week: a customer interaction, a technical decision, a number that surprised you. The bar for “good LinkedIn content” is lower than you think because most people aren’t posting anything at all.

The cringe fades. And unlike paid ads or PR agencies, it’s free. The only cost is your willingness to hit “post.”

That’s it for this week!

For more weekly dives into the world of early stage startups, subscribe below.

You are receiving this newsletter since you opted in earlier; if you would like to opt out of future newsletters, you can unsubscribe immediately.

This newsletter is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. This newsletter may link to other websites and certain information contained herein has been obtained from third-party sources. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation.

References to any companies, securities, or digital assets are for illustrative purposes only and do not constitute an investment recommendation or offer to provide investment advisory services. Any references to companies are for illustrative purposes only; please see a16z.com/investments. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund which should be read in their entirety.) Past performance is not indicative of future results.

Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Content in this newsletter speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see a16z.com/disclosures for additional important details.

Liquid Death is a healthy beverage company that looks like a beer brand, markets like an entertainment company, and grows like a tech startup. Founded in 2019 by Mike Cessario, a former ad creative with a background in making funny videos on no budget, the company has expanded from water to sparkling water, iced tea, and now energy drinks, doing hundreds of millions in revenue along the way. The company ran a Super Bowl ad this year which they produced internally for $300,000.

In a conversation with a16z speedrun, Cessario laid out the thinking behind the brand: why he assumes every product is a commodity, why comedy is the last unoccupied competitive moat in consumer goods, and what it actually takes to get a can of water onto a shelf at Kroger. Most of his advice applies far beyond beverages.

Watch the full conversation below, or scroll down for our top five takeaways from the conversation:

1. Comedy Is the One Thing Coke and Pepsi Can’t Do

Cessario’s career started in ad creative, and he’s always had a knack for making people laugh. He built Liquid Death’s competitive strategy around humor and entertainment in part because it’s structurally inaccessible to the biggest players in the market.

“Let’s say you want to do athlete marketing. Anybody with a checkbook can buy any athlete they want. If you’ve got the big enough number, you can get any snowboarder, any NFL athlete. So if you try to have a sports marketing theme for your brand, you’re just competing with every other giant brand that can just spend way more money than you can to, to, to buy more athletes or whatever. Comedy is the one thing that Coke and Pepsi can’t really do. Comedy is the one thing that Coke and Pepsi can’t really do. Even if they wanted to, true comedy can’t really survive in that bureaucracy.”

Cessario pointed to the experience of A-list comedians who’ve done deals with major brands, like Ben Stiller doing a Pepsi campaign. The story is always the same: everything funny got killed in the approval chain. “You actually had to work to make this person not funny,” he said.

2. Start by Assuming Your Product Is a Commodity

Cessario’s framework for thinking about brand starts from a premise that most founders resist: your product is probably not that different from your competitors’. He argued that founders get so deep in the weeds that they overweight tiny product differences that the broader market doesn’t notice or care about.

“Does Nike ever talk about how its materials last twenty days longer than Adidas? Any mustard on the shelf is probably all the same f***ing mustard. Can anybody pick out their favorite beer in a blind taste test? Probably not.”

His advice: start from the assumption that product differences are slim to none, then figure out how you win. Either you win on price and become the cheapest option, or you win because your commodity is far more memorable than everyone else’s. This thought exercise, he argued, is what forces you to invest in the things that actually create differentiation: brand, entertainment, and emotional connection.

3. An Entertainment Company That Monetizes via Beverage

Cessario described Liquid Death’s model inspired by the Red Bull playbook, with one substitution: comedy instead of action sports. The ethos is entertain people first, sell beverages second. That’s not a tagline. It’s the operational reality of how the company allocates resources.

“I made a video for fifteen hundred bucks that got three million views and generated more page followers than Aquafina with almost no investment. That’s proving out the model: if you can entertain people in commoditized categories, you’re going to have this huge competitive advantage because no one has any real product advantage.”

The economics have held at scale. Liquid Death’s Super Bowl ad cost $300,000 to produce internally, much lower than industry averages.

“We’re trying to make people laugh on the internet. As long as you can do that, you’re going to attract people and they have a reason to tune in. We’re occupying a place in their mind when they have five seconds to choose what to buy walking down the aisle.”

4. Cool Company, Not Just Cool Product

Asked to define what brand actually means, Cessario drew a distinction most founders miss. A lot of products are cool. Very few companies are cool. The strongest brands are the ones where people think the company itself is cool, independent of any specific product.

“There’s a lot of really cool products out there. But there’s very few cool companies. If you ask people, ‘What’s a really cool company?’ That’s where you start getting down to the people who are really good at brand.”

He used ChatGPT as a current example. Every LLM does more or less the same thing. But one brand dominates because it occupies a place in people’s heads, the way Google did with search.

Cessario also pushed back on the idea that mission alone drives loyalty. He told the story of an Uber driver who fell in love with Liquid Death’s “death to plastic” message. But the real reason she’d keep buying, he argued, is that the can gives her permission to participate in a counterculture she wouldn’t otherwise access. The mission is how she justifies it. “If that were really the only reason, the minute there was canned Smartwater, she’d buy that because it was cheaper.”

5. Distribution Is the Real Challenge

For founders who’ve only operated in software, Cessario’s description of beverage distribution is a wake-up call. Getting authorized in a retail chain is relatively straightforward. One meeting with a Walmart buyer can open four thousand doors. The problem is what happens next.

“As a small brand, you’re authorized in two thousand Krogers, but you’re only on the shelf in seven hundred, because the distributors just aren’t actually getting the product to the retailer. They don’t care, and there’s nothing you can do about it.”

The beverage industry runs on DSD (direct store delivery) networks owned by Coke, Pepsi, and the beer companies. You can’t use Coke’s or Pepsi’s distribution unless they own you. The beer networks are different because of the three-tier system: alcohol laws prevent producers from owning more than twenty percent of their distribution. So hundreds of independent distributors exist, and a new brand has to win them over one by one.

But even then, it’s a challenge to get distributors to get you priority:

“As a new brand, you’ve got to go one by one to all these guys and convince them to distribute your product ,to take your product and sell it to Walmart and Target and Kroger and 7-Eleven and everywhere else. But they’re also selling Budweiser and Bud Light and Michelob Ultra and all these other brands. So when you go in as a small brand, you’re so far down the totem pole. The rep is going into a Kroger and he’s only going to be there an hour. He needs to go to the next store. Most of his time is going to be restocking Bud Light.”

This, says Cessario, is why managing distribution is far and away the hardest part of scaling a new ready-to-drink beverage company.

Thanks to Mike Cessario for sharing his insights with us. And for more weekly dives into the world of early stage startups, subscribe below.

You are receiving this newsletter since you opted in earlier; if you would like to opt out of future newsletters, you can unsubscribe immediately.

This newsletter is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. This newsletter may link to other websites and certain information contained herein has been obtained from third-party sources. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation.

References to any companies, securities, or digital assets are for illustrative purposes only and do not constitute an investment recommendation or offer to provide investment advisory services. Any references to companies are for illustrative purposes only; please see a16z.com/investments. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund which should be read in their entirety.) Past performance is not indicative of future results.

Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Content in this newsletter speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see a16z.com/disclosures for additional important details.

🏰 Jon Lai proposes the “minimum viable moat” as the way to win in AI

🤖 Andrew Chen has a simple rubric for telling the difference between AI-native products and products with AI bolted on as an afterthought

💰 Emily Bennett has advice on fundraising: don’t lead with your “dream investor.”

💼 Join our talent network for more opportunities. If we see a fit for you, we’ll intro you to relevant founders in the portfolio

Somewhere around 2003, in a room full of investors, a video game crashed.

The game was called Republic: The Revolution, and it was supposed to be a generation-defining political simulation. Its creator, a young Londoner named Demis Hassabis, had pitched it with the kind of total conviction that makes people open their checkbooks. The demo failed. The product barely functioned. But Hassabis forged ahead, telling investors a story about what the game would become.

Ultimately, the game shipped to middling reviews, and Hassabis’s company, Elixir Studios, folded. Hassabis went quiet for a few years, then started a small research lab called DeepMind. Everyone in Silicon Valley knows the story from here. Google acquired DeepMind and the company has gone on to develop technologies that power Google’s frontier AI models, and which have earned Hassabis the Nobel Prize in Chemistry.

Sebastian Mallaby has been writing about entrepreneurs like Hassabis for twenty years. His book More Money Than God went inside hedge funds, The Power Law chronicled venture capital, and his new book, The Infinity Machine, tells the story of Hassabis and, through him, the story of how modern AI came to exist.

We spoke with Mallaby about what he’s observed across all three of these worlds, and why the Hassabis story might be more useful to founders than they realize.

Conviction Vs Delusion

Hedge funds, startups, venture capital: in each case, a small group of people has decided that everyone else has it wrong, and is betting real money on that belief. These are all fields where the gap between conviction and delusion is essentially invisible from the outside. The only thing that separates them is the outcome.

We asked Mallaby: Is this tension between conviction and delusion the real throughline connecting your work?

“I think that’s a fair characterization of the whole investment boutique entrepreneurship ecosystem,” Mallaby said. “In hedge funds, you’ve got market prices which you think you can outguess, and those market prices are the weighted average of intelligent people’s judgment. So you need to be quite bold to think you know better. In the same way, a startup entrepreneur is trying to disrupt incumbents who are not idiots. You have to be bold.”

But the line between conviction and delusion can at times appear thin. And as Mallaby tells it, even the most successful founders have found themselves on both sides of that line.

The Fluency Test

One of the early stories told in The Infinity Machine recounts how the dev team on Republic: The Revolution slipped into a pattern of delusional belief about the game’s chances at success. “The engineers knew he [Hassabis] wouldn’t accept anything short of ‘yes, we can do this,’” Mallaby told us. “So they said that. And that made Hassabis even more determined not to accept no for an answer. Because he was hearing yes from his own team, he got even more confident. There was this feedback loop.”

What matters about this story is what happened after. Mallaby frames the Elixir collapse as the event that made DeepMind possible.

Years later, facing a similar crossroads with AlphaFold, Hassabis found himself in the position of deciding whether to push a team past the point where its own leader said the goal was impossible. The head of the AlphaFold effort, Andrew Senior, told him flatly in 2018: we’ve built the best protein-folding system in the world, but we cannot do what you’re asking us to do. Instead of overriding him, Hassabis began sitting in on meetings with the broader team and listening closely for what Mallaby calls “the fluency of the conversation.”

“If lots of ideas are coming up, he figures, well, it’s not like they’re stuck,” Mallaby says. “If they had more time or resources, they could test out some of these ideas they’re kicking around in the conference room. Fluency was his metric for whether to go ahead.”

A room full of people generating ideas looks fundamentally different from a room that’s gone quiet. The first means there’s potential. The latter means the problem might be a dead end. The fluency test was how Hassabis learned to tell the difference, and it was a lesson hard-won by a founder who’d experienced the downsides of charging ahead when the room was telling him what he wanted to hear.

Mallaby put the takeaway plainly: “I think it’ll resonate with a venture capital firm that second-time entrepreneurs are often the ones you most want to back.”

On Communicating Conviction

For many practical and structural reasons, the dominant way that Silicon Valley founders talk is prospective. You announce what you’re going to build before you’ve built it. You generate excitement around a future state that may or may not arrive. Some of the most visible figures in tech have operated this way for years, tweeting about ambitious new capabilities before the technology can deliver them. It works as a fundraising mechanism. It also, as Mallaby noted dryly, sometimes involves announcing things that aren’t yet certain to work.

Hassabis, at least in his mature period, worked in the opposite direction. After DeepMind’s Atari system started producing real results in 2013, his instinct was to take the concrete achievement and wring every possible drop of credibility out of it. The target he picked was Nature magazine. “Nobody else on his team thought that publishing a paper in Nature was even conceivable,” Mallaby said. “Nature didn’t really do computer science. But Hassabis cultivated an editor at Nature for a year, then got him to agree to look at a submission from DeepMind, then got the DeepMind video team to design cover art so Nature would put it on the front.”

These kinds of details accumulate throughout the book. Hassabis comes across as someone who wastes very little, who finds a use for whatever’s already lying around, and who treats reputation as a resource to be deployed strategically rather than as a mere byproduct of the work.

“He was absolutely as focused on communication as anyone else in AI,” Mallaby said. “But he had this retrospective approach, of playing up actual achievements, as opposed to announcing future ones.”

On Timing and “Scientific Taste”

Marc Andreessen likes to say that one of the main reasons great entrepreneurs fail isn’t because their idea is wrong, but because they were simply too early. Predicting when the market is ready for something is one of the hardest jobs of the entrepreneur, and a thread running through Mallaby’s book is that Hassabis foresaw a lot of things before others, including the possibilities of reinforcement learning and the possibility of delivering something meaningful for the protein folding problem.

Mallaby told us that Hassabis uses the phrase “scientific taste” to describe the ability to sense when a field is about to tip, even when most observers still see stagnation. It’s a phrase that does a lot of quiet work, treating taste as pattern recognition operating at the boundary between what the data supports and what it merely suggests: the ability to read straws in the wind when everyone else is staring at the ground.

“With DeepMind’s Atari agent, which is an early experiment in combining deep learning with reinforcement learning, Hassabis sees when something is poised to advance,” Mallaby said. “And then he repeats the same trick with the decision to go after AlphaGo, even though Sergey Brin had told him it’s crazy. After Atari he feels you can do Go. Later, after Go, he feels you can do protein folding.”

David Silver, one of DeepMind’s key researchers, gave Mallaby a metaphor for this: Hassabis had understood “the ladder.” Each project was grander than the last, and each one funded the next: financially, and in terms of credibility. To get to the moon, the DeepMind team figured out which rung to stand on next.

Internally, the DeepMind team ran two modes at once: a cheap exploration layer, where researchers pursued open-ended questions, and an expensive exploitation layer, where strike teams poured everything into a single target. Hassabis’s job, as Mallaby described it, was knowing when to shift from one to the other. “He’s willing to have people exploring different avenues for AI, doing their thing,” Mallaby said. “But then he’s going to have the scientific taste to know which one to push on. And when he’s pushing, it’s a totally different organizational structure.”

How to Predict the Future

We asked Mallaby whether, after spending years embedded in this story, he felt he understood where AI was actually going. He gave a modest yes.

“My view is, if you look at what’s happened since ChatGPT came out, the number of major jumps that the technology has made is flabbergasting,” he said. “You go from something which basically hallucinated nonstop to something that pretty much got that under control. Longer context windows. Multimodal. Reasoning. Now it’s starting to be agentic. Next is world models. And it’s only been three and a half years.”

Every time someone declares a ceiling, Mallaby pointed out, a new axis of scaling appears to replace the one that supposedly ran out. He’s bullish on infrastructure. The application layer, he admitted, is anyone’s guess.

But the question we kept circling back to was the one about conviction and delusion, and whether there was any reliable way to tell them apart before the outcome settles the matter. Mallaby’s answer, when it finally came, had the weight of someone who has spent two decades in rooms with people who stake everything on their own judgment.

“You don’t necessarily get the balance correct the first time,” he said. “But the only way you learn is by doing.”

The Elixir game crashed, and the company died, and then the same person started something else.

Thanks to Sebastian Mallaby for speaking with us. And for more weekly dives into the world of early stage startups, subscribe below.

You are receiving this newsletter since you opted in earlier; if you would like to opt out of future newsletters, you can unsubscribe immediately.

This newsletter is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. This newsletter may link to other websites and certain information contained herein has been obtained from third-party sources. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation.

References to any companies, securities, or digital assets are for illustrative purposes only and do not constitute an investment recommendation or offer to provide investment advisory services. Any references to companies are for illustrative purposes only; please see a16z.com/investments. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund which should be read in their entirety.) Past performance is not indicative of future results.

Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Content in this newsletter speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see a16z.com/disclosures for additional important details.

📈 We’re hiring a Marketing Partner to support our new Alpha fellowship program! See Katia Ameri’s post here for details.

🚀 As SR006 Demo Day approaches more founders are coming out of stealth like Audrey Lo, who’s building AI caretakers for seniors; Yusan Lin, who’s building a marketplace for fashion models to license their likenesses; and Ruming Zhen, who’s building AI agents for real-world context.

🔎 Bella Nazzari says using Claude to mass-apply to jobs isn’t likely to work. But she has a method that will.

💼 Join our talent network for more opportunities. If we see a fit for you, we’ll intro you to relevant founders in the portfolio

Rahul Vohra learned how to make things fun before he learned how to make them useful. He started his career as a game designer on RuneScape, created the Gmail plugin Rapportive (acquired by LinkedIn), and then spent three years building Superhuman before letting the public anywhere near it. Along the way, he developed a widely shared framework for measuring product-market fit.

In a fireside chat with a16z speedrun Visiting Partner Fareed Mosavat, Vohra laid out the playbook behind Superhuman’s rise. We discussed why gamification fails but game design works, why he turned away customers who wanted to pay him, and why the right price for email is thirty dollars a month.

Watch the full conversation below, or scroll down for our top five takeaways from the conversation.

1. The Product Market Fit Engine

By April 2017, Superhuman was two or three years into development and the team was pushing hard to launch. Vohra suspected the product would flop. But “we don’t have product-market fit, keep working” is not an inspiring message. So he built a system to fix it and rally the team around the process.

The foundation was a question from growth pioneer Sean Ellis: “How would you feel if you could no longer use the product?” Users answer one of “not disappointed, somewhat disappointed, or very disappointed.” Ellis benchmarked hundreds of startups and found that companies reaching product-market fit almost always had more than 40% say they’d be “very disappointed” to lose access to the product. Companies that didn’t almost always had less.

Vohra decided to apply this framework to Superhuman:

“I ran the first question and we had twenty-two percent very disappointed. Way less than forty. You might think I was bummed out. Actually, I was excited. Finally, I could explain what I was feeling to the team in a way that was irreproachable. You can’t argue with these metrics.”

Vohra’s added three questions of his own: Who is the product best for? What’s the main benefit? How can we improve it? The first move was changing the market, not the product. By dropping non-core personas (data science, sales, engineering) and focusing on VCs, CEOs, founders, and BD, the score jumped from twenty-two to thirty-two percent with zero product changes.

Then he focused on the “somewhat disappointed” users for whom the main benefit already resonated. Only a few things were holding them back. He worked down that list systematically, splitting energy between building more of what users loved and fixing the boring complaints. Vohra reports on the results:

“The quarter after, we were at forty-six percent. The quarter after that, fifty-six percent. The quarter after that, fifty-eight percent.”

2. You Have the Right to Deny Service

Superhuman didn’t have an Android app for years. It lacked an Outlook integration and a Salesforce connector. Rather than let those gaps create detractors, Vohra simply didn’t let those people in. He screened users with a survey before they could sign up.

“We’d be like, what mobile phone do you have? They’d say Android. And I said, well, you can’t have Superhuman. And people would get confused. Like, why isn’t it my decision whether or not I buy? And I’m like, no, it’s my decision whether or not I sell.”

The logic was straightforward: every user who has a bad experience is a net detractor. Controlling who gets in means controlling the narrative around your product. Vohra’s goal was to make sure every single person who used Superhuman became a net promoter. That meant being willing to say no to people who wanted to give him money.

3. Game Design vs. Gamification

Vohra started his career as a game designer on RuneScape, and he brought those instincts directly into Superhuman. But he drew a sharp line between game design and gamification. Points, badges, streaks, and leaderboards can push behavior. They don’t make the underlying thing fun.

He cited a Stanford study from the 1970s. Researchers divided kids into two groups: one promised a reward for drawing, the other given no incentive. Over the following days, the rewarded kids spent half as much time drawing. The extrinsic motivator killed the intrinsic motivation.

“That’s exactly what happens with gamification. Giving your users a reward for something can reduce their intrinsic motivation to do that thing. To build products that feel like a game, you have to go into the theory of game design: what is a game, what is play, what is fun?”

His approach centers on building “toys” within the product. A toy invites playful exploration. Superhuman’s time autocompleter is his go-to example: users type “2D” for two days, “3MO” for three months, then start trying to break it. Someone typed “never.” The product responded with a shrug emoji. These localized moments of delight, built on what Vohra calls “squishy affordances,” are the foundation of a product that people genuinely enjoy using.

4. The Human-Touch Onboarding Machine

For years, Vohra personally onboarded every new Superhuman user. Five people a week, two-hour sessions, in person at their office. He’d research them beforehand, bring a gift (whiskey, wine, or tea), watch them use Gmail, demo Superhuman, ask for their credit card, then observe them using the product for as long as they’d let him.

“If you haven’t done this, please do it. And please never stop doing it. Even with a mature product, it is incredible how much you will see that maybe even your product engineers or your PMs won’t see. You as the founder are going to see every single problem.”

The team pushed him to grow faster. He refused until every bug from those sessions was fixed. “Every user has bug fatigue. They might tell you one or two or three bugs. They’re not going to tell you ten or twenty or thirty. And we definitely had that many.”

Then the metrics came in: activation, retention, virality, and NPS were all off the charts. Vohra tested whether it was him or the process by having his brother (then head of growth) run the onboardings. His brother was even better, so it was the process. That’s when Vohra hired onboarding specialists in pairs, scaled to forty Zoom calls a week, and turned it into a machine. Investors saw this approach as a genuine strategic edge.

5. Premium Pricing and the Prosumer

Superhuman launched at $30 per month for email. Vohra deliberately onboarded the five most influential people he could find each week, and he priced the product to match how those people think about themselves.

“Email is kind of part of the ego. It’s like, ‘Oh yeah, I get this many emails a week.’ The pressure is part of the ego. And I was like, well, if that’s the case, you know what’s going to make you feel good about it? Spending a lot of money on fixing that problem.”

He framed the broader strategy as the “prosumerization of the enterprise.” A prosumer, in his definition, is a power user with all the economic agency but none of the time. They have the money to spend on solving a problem; they just don’t have the patience to figure things out. That’s the customer Superhuman was built for.

His one regret on pricing: not raising it sooner. Superhuman stayed at $30 per month for years before adding a $40 tier. Vohra estimated they left 30% of ARR on the table. His tactical advice to founders: raise prices every year to stay ahead of inflation.

Thanks to Rahul Vohra for sharing his insights with our founders. And for more weekly dives into the world of early stage startups, subscribe below.

You are receiving this newsletter since you opted in earlier; if you would like to opt out of future newsletters, you can unsubscribe immediately.

This newsletter is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. This newsletter may link to other websites and certain information contained herein has been obtained from third-party sources. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation.

References to any companies, securities, or digital assets are for illustrative purposes only and do not constitute an investment recommendation or offer to provide investment advisory services. Any references to companies are for illustrative purposes only; please see a16z.com/investments. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund which should be read in their entirety.) Past performance is not indicative of future results.

Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Content in this newsletter speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see a16z.com/disclosures for additional important details.

We're gearing up for New York Tech Week (June 1-7), and we want you to host an event. If you've been wanting to throw a founder dinner, a demo day, a hiring happy hour, or really anything that gets great people in a room together, this is your moment.

Sign up to host here.

This week’s essay is a guest post from a16z speedrun investor Marcus Segal. He draws on his decades of startup experience to offer some time-tested advice on seeking, and keeping, product-market fit.

Emmett Shear, the co-founder of Twitch, visited speedrun founders for a session recently. Someone asked him how he would describe product-market fit. His answer, “How do you describe being in love? You know it when you’re in love.”

Fair enough. But the job of a founder isn’t to sit around waiting to fall in love. The job is to put your best foot forward and to avoid looking for love in all the wrong places. And the combination of a team of engineers cranking out features with AI, a lack of vision for your product and an undisciplined feature list is exactly that—looking for love everywhere all at once with a bad haircut!

When I was at Zynga I was fortunate enough to work closely with Mark Pincus—a founder that had never shipped a game—to build one of the biggest gaming companies on Earth with 300M players. Along the way I’ve co-founded two other startups in B2B and B2C and had the opportunity to support and learn from hundreds of founders on their quests for PMF while mentoring at Y Combinator, Google Launchpad, and now as an investor on the a16z speedrun team.

Across all of these experiences and through massive disruptions in technology three things have never changed:

1. A CEO’s job is to not run out of money before finding PMF

2. Engineering person hours are the most important resource in your company

3. Finding and keeping PMF is more likely to occur with a roadmapping process

It is my hope to share what I have seen work at Zynga and countless other startups in the hopes that all of you can find PMF.

PMF Is a Journey, Not a Destination

When I joined Zynga, none of us really knew how to make great games. What Mark Pincus understood, though, was how to build a system for finding what works—fast.

Our competitors released content once every two weeks, in English. We released five to ten content updates per week, per game, in 26 languages. We didn’t have AI. We generated an enormous number of hypotheses that could be rapidly tested and threw an army of PM’s,engineers, artists and LOTS of QA onto anything that found heat. The work of a 40 or 50 person team at Zynga could now be done with 10 people.

The methodology we used at Zynga wasn’t complicated. The best processes rarely are. Here’s the version I recommend at speedrun—adapted for teams that are building with AI tools to iterate on live products to deliver results faster and better.

Start with the metric, then build the feature.

Before you create a single line of code for a new feature, answer one question: what number are you trying to move? Reach? Retention? Revenue? The specific metric doesn’t matter as much as the discipline of choosing one. Because if you can’t articulate which number a feature is supposed to move, you have no way of knowing whether it worked or was accretive to the product experience. It should be one sentence, “We will launch Product X feature to give Y value to the customer moving Z metric.”

Get all the ideas out, then stack rank them.

This is where most founders stall or face major coordination breakdowns on their teams. All the builders have a running list of features in their heads and they build whatever feels most urgent on any given day. Don’t do that. Start by getting every idea out of your heads and onto a list. Then rank them as a team by asking a simple question: which of these has the highest probability of moving the metric we care about?

T-shirt size the effort.

Vibe coding has broken our intuitions about how long things take to build, particularly across teams.. Pre-AI, experienced engineers got pretty good at estimating effort. Now the volume of code generated across a team of engineers creates new layers of complexity in coordination and design (to name a few).

Things that seemed trivial take forever to polish, and things that seemed hard take twenty minutes and merging everything into a thoughtfully designed UI/UX framework is tougher than ever. And let’s not forget We all need to build or rebuild this muscle for the AI era so estimate and track actuals. It’s a new skill so expect a lot of learning along the way.

Power rank by impact-to-effort.

Now you have a list of features, ranked by expected impact, with effort estimates attached. The math practically does itself. High impact, low effort? Do it first. Low impact, high effort? Maybe never.

Roadmap looking one-week-back, one-week-forwards view.

At Zynga, we looked one week back and three weeks forward every week in a team review. What did we build last week? What did we think would happen? What actually happened? The metrics provided the answers. MVP’s of features were tested and scrapped if they did not perform. Course corrections happened to the roadmap to avoid wasting engineering time. Learning compounded quickly and spread across all of our games.

With today’s tools that cadence would be way too slow. Today we look one-week-back and one week forward. The cadence matters. It’s short enough to course-correct fast, but long enough to actually ship meaningful work.

Measure everything.

At every successful startup or scale up I have seen up close the winners are obsessed with metrics. Great ideas can come from any person in your organization.If someone on the team pitches or builds a feature and accurately predicts what it will do they should be lauded. If they are wrong, nobody gets fired. The next time, they will calibrate better. That’s the culture you need. Experience is never wasted if your team learns and gets better.

The Gift

Here’s what I want founders to take away from this: I believe a deliberate roadmapping process is the highest-leverage activity available to you as a founder to increase your chances of finding and keeping PMF. You will maximize your most valuable resource engineering time and grow a killer team in the process.

Enjoy the journey.

Got questions for Marcus? Let us know in the comments. And for more weekly dives into the world of early stage startups, subscribe below.

You are receiving this newsletter since you opted in earlier; if you would like to opt out of future newsletters, you can unsubscribe immediately.

This newsletter is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. This newsletter may link to other websites and certain information contained herein has been obtained from third-party sources. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation.

References to any companies, securities, or digital assets are for illustrative purposes only and do not constitute an investment recommendation or offer to provide investment advisory services. Any references to companies are for illustrative purposes only; please see a16z.com/investments. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund which should be read in their entirety.) Past performance is not indicative of future results.

Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Content in this newsletter speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see a16z.com/disclosures for additional important details.

Guillermo Rauch is the founder and CEO of the web development platform Vercel and the creator of Next.js, one of the most widely used frameworks on the internet.

In a fireside chat with a16z Investment Partner Gabriel Vasquez, Rauch explained how open source can accelerate product-market fit and why ruthless focus matters more than ever in the age of AI.

Watch the full fireside below, or scroll down for our top five takeaways from the conversation.

1. Open Source Is a Speedrun to Product Market Fit

Rauch’s framework for open source is to treat it as a filter. The two hardest problems for any startup are getting people to know you exist and building the right thing. Open source, in his view, stress-tests both at once.

“I’ve always thought of open source as a speedrun to product-market fit, because if people don’t use it when it’s free, when it’s easy to consume, when the code is available, then you probably should be working on something else.”

He acknowledged the obvious downside: openness makes replication easier, especially in the age of agents. But Rauch argued that the best open source projects extend into broader platforms, and success surfaces new problems to solve. Next.js itself is only a sliver of the cloud problem enterprises actually face.

“What stops entrepreneurs in their tracks is that they worry too much about what’s going to happen in the future. In reality, they don’t even have product-market fit yet.”

2. Unbounded Vision, Ruthless Focus

Asked about the most important focus decision in Vercel’s early days, Rauch described a tension he calls the “bicameral mind.” You need a bold, boundless long-term vision. But on any given Tuesday with three people in the room, that same ambition will destroy you if you let it dictate what you build.

“When you’re getting started and you’re three people, if all you do every day is unbounded ambition and building two hundred million different features, then you’ve fucked up.”

Early Vercel tried to make everything deployable: every language, every runtime, every framework, even databases. Then Rauch forced the question: where are we being truly excellent? The answer was the front-end layer, a new kind of managed service that didn’t exist yet. The result was the difference between building something no one would pay for and building something worth hundreds of millions.

“If I had continued doing everything, I would have made a company that made zero dollars. Literally zero dollars. Discerning buyers want the best of each class of problem. They don’t settle for the second-best solution.”

3. Choosing What to Build Matters Most

In an era where AI makes everything buildable, the temptation is to build everything. But that’s exactly backward. When every problem becomes shallow with enough tokens, choosing which problems deserve your attention becomes the real competitive edge.

“A bunch of you are probably already over-building massively because you notice that you can just prompt it and it’ll do it. But the discernment of what problems are worth solving could be one of your biggest differentiators.”

He illustrated the point with a contrast from inside Vercel. A good use of AI: having an agent work overnight on a JavaScript optimization that made the platform twenty to forty percent faster for customers. A bad use: rewriting your internal HR tools. The CEO’s job, Rauch argued, is increasingly about helping the company decide what’s worth investing in when everything is possible.

4. Hire for Proof of Work and How People Package Their Ideas

On talent, Rauch’s advice was to look for people on the rise rather than people who’ve already arrived. Find the ones shipping high-quality work, especially with AI tools, and pay attention to something most hiring managers ignore: how candidates communicate what they’ve built.

“Sometimes the reason I reach out to people is not just what they built, but how they communicate it.”

He told the story of sliding into a developer’s DMs after seeing their novel approach to a well-known UI problem, while also finding a new way to package and explain a contribution. Rauch extended this into a broader product philosophy, comparing it to Apple’s unboxing experience: the website, the store, the feeling of lifting the lid. Every layer of the experience has been thought through.

“I sometimes see slop landing pages from startups. It’s a recording of the entire desktop of the entrepreneur, and it’s just so busy. I see his dock with two hundred applications, his entire browsing history, two hundred plugins. Delete all the things that are not your thing, because they’re an excuse for people to get distracted.”

5. Choosing What to Dismiss Is a Competitive Advantage

Rauch closed with a point about conviction. In a world where every new tool and framework generates immediate hype, the ability to say “no” is underrated. His approach is to never outsource judgment. He downloads products, signs up, uses them firsthand. If he can’t deeply understand something himself, he doesn’t build conviction around it.

“Silicon Valley has this sharp edge of people just get so f***ing excited about everything. And some of those things are worth dismissing. Choosing what you dismiss can also be a competitive advantage.”

The two-phase process he described: experiment broadly, then set a high bar for what you actually bring to market. When you do find something that clears that bar, go all in. Post it everywhere, email it, present it. But don’t confuse the excitement of experimentation with the conviction required to ship.

“If I don’t fully understand it myself, it’s very hard for me to actually build conviction.”

Thanks to Guillermo Rauch for sharing his insights with our founders. And for more weekly dives into the world of early stage startups, subscribe below.

You are receiving this newsletter since you opted in earlier; if you would like to opt out of future newsletters, you can unsubscribe immediately.

This newsletter is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. This newsletter may link to other websites and certain information contained herein has been obtained from third-party sources. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation.

References to any companies, securities, or digital assets are for illustrative purposes only and do not constitute an investment recommendation or offer to provide investment advisory services. Any references to companies are for illustrative purposes only; please see a16z.com/investments. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund which should be read in their entirety.) Past performance is not indicative of future results.

Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Content in this newsletter speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see a16z.com/disclosures for additional important details.

What will happen to spreadsheets in the age of AI?

Already, a transition is underway. The big AI labs are building their own integrations for agents in Excel and Google sheets. Startups are getting in on the action too, with teams like Fundamental Research Labs (SR002) earning traction by reimagining the spreadsheet experience to be AI native.

Last week, a16z speedrun General Partner Andrew Chen kicked a hornet’s nest on X with a post arguing that “anything that is currently modeled as a spreadsheet is better modeled in code.”

Here’s the full post, to save you a clickthrough:

AI code gen means that anything that is currently modeled as a spreadsheet is better modeled in code. You get all the advantages of software—libraries, open source, AI, all the complexity and expressiveness.

Think about what spreadsheets actually are: they’re business logic that’s trapped in a grid. Pricing models, financial forecasts, inventory trackers, marketing attribution—these are all fundamentally *programs* that we’ve been writing in the worst possible IDE. No version control, no testing, no modularity. Just a fragile web of cell references that breaks when someone inserts a row.

The only reason spreadsheets won is that the barrier to writing real software was too high. A finance analyst could learn =VLOOKUP in an afternoon but couldn’t learn Python in a month. AI code gen flips that equation completely. Now the same analyst describes what they want in plain English, and gets a real application—with a database, a UI, error handling, the works. The marginal effort to go from “spreadsheet” to “software” just collapsed to near zero.

This is a massive unlock. There are ~1 billion spreadsheet users worldwide. Most of them are building janky software without realizing it. When even 10% of those use cases migrate to actual code, you get an explosion of new micro-applications that look nothing like traditional software. Internal tools that used to live in a shared Google Sheet now become real products. The “shadow IT” spreadsheet that runs half the company’s operations finally gets proper infrastructure.

The interesting second-order effect: the spreadsheet was the great equalizer that let non-technical people build things. AI code gen is the *next* great equalizer, but the ceiling is 100x higher. We’re about to see what happens when a billion knowledge workers can build real software.

The post racked up over a million views and hundreds of replies, with not everyone in agreement.

Richard Pham argued that much of the value of spreadsheets is in the grid-based visualization itself:

Allie Harris came at this question from a different angle:

Tom Robbins was skeptical as well, pointing to the continued ubiquity of PDF files:

This was a a comparison Andrew was happy to run with:

The question is whether spreadsheets will follow the same trajectory—still around, but no longer the default way a billion people interact with data and business logic.

An interesting framing in the thread came from Dan Hockenmaier, who pointed out that there are really two kinds of spreadsheet use. The “mini software” kind—dashboards, trackers, attribution models—is almost certainly getting replaced. But the other kind where you develop a financial model you build to hone your own understanding of a business is stickier.

One top reply came from Robert L. Peters, who offered a full-throated defense of the spreadsheet:

Would bet my net worth with 100x leverage that front office finance jobs will still use spreadsheets 10 years from now

Can you run an LBO model with python, audit each cell (F2), make micro adjustments to formulas with instant feedback, toggling an assumption back and forth to see how it impacts a sensitivity table that is 1000 calculations later in the chain? And do this with someone who just got out of college a couple months ago?

The act of making the spreadsheet in many cases forces the user to think about assumptions, and reason about the drivers. You may not always have it fully formed in your head in advance as a prompt until you see the data, create an operating build, etc.

If an MD draws on their iPad “make a box here, increase spacing in these three areas, change the calculation for this one year of EBITDA to xyz, and make an italics EBITDAR line underneath, break out product lines in revenue build, shift to monthly, etc” they can give to an analyst and can get back exactly what they want to see in an hour. They don’t need to wait 12 months for a SaaS platform to maybe add the feature. Or prompt an LLM that may asymptote on what you want but isn’t perfect.

There are also so many edge cases for bad data from private companies where you really need to manually look at the data points to understand the gaps. Python doesn’t allow you to do that and prompting doesn’t provide a visual surface.

I’m sure you will be able to prompt a lot of this stuff but for many use cases the form factor will continue to be a spreadsheet - the analyst will be able to make tweaks to the spreadsheet output the LLM makes, check the logic easily, make edits where it’s faster to make it the way you want vs type out prompts, etc

In front office finance jobs (maybe other than VC which is less precise with modeling), one single calculation error can dramatically change the answer and cost you $1 million or $1 billion. 95% right is 0% useful. Python and prompting do not allow for full verification of the calcs and data through to the primary sources.

Tech ppl perennially have a fundamentally flawed mental model of Excel (“why not use Python????”) because they have never had exposure to how it’s used in the real world, and it’s why so many companies with the business model “move your OUTDATED spreadsheets to my MODERN B2B SaaS in your browser!” that for the past decade have spammed PE firm inboxes have crashed and burned

Spreadsheets are a better form factor than python for business users and LLMs won’t kill spreadsheets

Peters’ reply gets at something the original take arguably doesn’t address: The act of building the model is how you develop conviction about the inputs. You can’t just skip to the output.

Chen pushed back:

What you’re talking about is the “IDE” user experience of spreadsheets -- being able to inspect the logic, checking out values/cells/etc, debugging, etc. I both agree that you’ll need to do that BUT also that this is all possible, and frankly, better, in software as well

In software you also get unit/integration tests, version control, coding harnesses, AI everything, and much more. You could even ask the LLM to describe the code in a grid if you want to try to be quasi-backwards compatible.

Chen is saying those workflows migrate to a better substrate.

But you can’t win ‘em all. Joel Grus came off the top rope with this zinger:

So, after a day of fighting in the reply guy trenches, Chen posted a followup (full text below):

Reading the replies -- a lotta folks hate this prediction!

A lot of folks can’t imagine programming the logic/variables/inputs without the spreadsheet grid paradigm

My primary counterpoint:

• Much complaining is from people trained on keyboard shortcuts in Windows Excel on Thinkpads during their glorious banking days, swearing that they’ll never switch to anything else. Late adopter normie finance bros. Soon to be disrupted

My actual counterpoints:

• Programming itself has changed its UX many many times. Punchcards, typing into files, IDEs, and now LLM coding harnesses. Spreadsheets are not the only way to encode business logic -- there are better ways, while gaining all the power of software

• The grid UX might remain in some form, but might be more of a display. Just as you code in Codex/Claude but then still pull up a webpage. Or maybe you’ll have a grid as a DB but then build apps on top, but still want a querying UI for the data

• LLMs will make going between logic in code and logic in spreadsheets interchangeable. So maybe you’ll edit in a grid but then hit “deploy” and it’ll build a webapp in the cloud. And just as we have VLOOKUP() they’ll be LLM() that can encode AI logic

• Anyone who works with software knows its infinitely better and more expressive and more powerful. AI code gen is a blessing for all the non-technical excel wizards who can now take their work to the next level

The history of computing really is a story of interfaces evolving beyond what the previous generation thought was possible. Punchcards felt essential until they didn’t. Command lines felt essential until GUIs arrived.

The spreadsheet grid might be next.

What do you think about the spreadsheets debate? Let us know in the comments. And for more weekly dives into the world of early stage startups, subscribe below.

You are receiving this newsletter since you opted in earlier; if you would like to opt out of future newsletters, you can unsubscribe immediately.

This newsletter is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. This newsletter may link to other websites and certain information contained herein has been obtained from third-party sources. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation.

References to any companies, securities, or digital assets are for illustrative purposes only and do not constitute an investment recommendation or offer to provide investment advisory services. Any references to companies are for illustrative purposes only; please see a16z.com/investments. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund which should be read in their entirety.) Past performance is not indicative of future results.

Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Content in this newsletter speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see a16z.com/disclosures for additional important details.

Ben Horowitz has been advising founders for decades.

In a recent fireside chat with a16z speedrun General Partner Jonathan Lai, Ben made the case that the fundamentals haven’t changed nearly as much as people think. Strategy is still about articulating a clear “why.” Product management is still about shipping the right product at the right time. And the best fundraising pitch is still the one you actually believe.

What has changed is the landscape around those fundamentals. And there are massive new opportunities hiding in plain sight, if you know where to look.

Watch the talk in full below, or read on for five takeaways that stood out to us.

1. Your Story Is Your Strategy. Write It Down.

Horowitz opened with a point that sounds simple but that most founders get wrong: your company’s story and its strategy are the same thing. Not adjacent. Not related. The same. And if you aren’t constantly rewriting that story as you learn, your strategy is drifting without anyone noticing.

“The story of the company is also the strategy of the company. Nobody’s got some strategy in their back pocket that’s entirely secret and then a story that’s different.”

He pushed founders to write it in long form, because that forces the most discipline. The story answers the “why”: Why does this company exist? Why should someone join? Why should someone invest? If people know the why, Horowitz argued, you don’t even have to tell them the what.

“The problem that founders have is it doesn’t feel like work. ‘What are you doing today?’ ‘Oh, I’m working on my story.’ That doesn’t sound like work. But it’s probably one of the most important parts of the job.”

2. In the AI Era, Hire for Creativity and Relationships

Asked what traits founders should prioritize when hiring, Horowitz pointed to two that Silicon Valley has historically undervalued: creativity and relationship-building. AI already handles the grind-it-out tasks. What it can’t replicate is someone who walks in with original ideas or who can build and maintain high-fidelity relationships.

“If you don’t have instincts around those things, I don’t know that all the AI in the world can help you. Whereas the grind-them-out tasks, AI is already really good at.”

He illustrated this with a story from his own firm. They had built a legendary marketing organization, but it was rooted in 2009-era traditional media. It wasn’t until they brought in people who were genuinely creative about new media that the firm could break through on that front. Original thinking, Horowitz noted, is rarer than people assume.

3. “Right Product, Right Time” Is Still the Only Job

Horowitz’s essay Good Product Manager, Bad Product Manager has become required reading in product circles. But the origin story is more blunt than most people realize. He wrote it because he was furious at seven PMs who were doing everything except the one thing he needed from them.

“I was so f***ing mad at them because I was like, ‘You people are doing everything but your job.’ You’re writing these requirements, you’re running around, pitching customers this and that. It’s a lot of work. But your job is: right product, right time. And if you don’t give me the right product at the right time, I don’t give a f*** what you do.”

When Lai asked whether the AI landscape changes the PM role, Horowitz was characteristically direct. The landscape doesn’t change the job. It’s still about “right product, right time.”

“You sound like my f***ing product managers. ‘Oh, now the landscape is different. I won’t write a PRD, I’ll write a prompt.’ Okay, great. Is it going to get you the right product, right time? If not, it doesn’t matter… And if you can’t do that, then Opus 4.6 can do all those tasks. No problem.”

---

4. Build Where There Are Shortages

Asked about the most exciting opportunities for founders, Horowitz pointed to areas where there are shortages. If there’s unlimited demand for AI tokens, the interesting question becomes: what supply-side bottlenecks can you solve?

“We’ve got an electricity shortage, chip shortage, token shortage. We’re going to have a cooling shortage. There’s all kinds of levers that you can pull to alleviate some of those shortages, and those will be valuable.”

There are examples Ben pointed to, like a16z’s decision to back Heron Power, a company building power transformers, because the shortage is so severe that it’s a bottleneck for new power plant construction. Horowitz also flagged a second category: long-standing problems of humanity, like cancer, that could become solvable thanks to new technology.

5. The Best Fundraising Pitch Convinces You First

When the conversation turned to fundraising, Horowitz’s advice cut against the instinct most founders have when walking into a pitch meeting: trying to figure out what the investor wants to hear. Instead, you should be able to convince yourself first:

“Being able to articulate why you’re building this company in a way that convinces even you that you should re-join your company is really the thing. If you can convince yourself completely that this is the greatest freaking idea in the world, then that’s the most compelling thing… The mistake people make in fundraising is they go, ‘Well, what does Ben want here? Does he want to see a graph like this? Does he want to know this about market size?’ That’s very hard to figure out, and you usually get it wrong. But you can know: how would I convince myself to join this company, or invest in this company? That always works the best.”

The reason is practical, not philosophical. A pitch built on what you believe is a pitch you can’t get talked out of. The worst outcome in a fundraising meeting is someone poking a hole in your narrative and you folding, because the argument was never yours to begin with.

“It’s kind of like a bad way to start a relationship. ‘I’m going to pretend to be this other person and then you’re going to like me, but then we’re going to get together and you’re going to find out the real me.’ That tends not to work out either.”

Thanks to Ben Horowitz for sharing his insights with our founders. And for more weekly dives into the world of early stage startups, subscribe below.

You are receiving this newsletter since you opted in earlier; if you would like to opt out of future newsletters, you can unsubscribe immediately.

This newsletter is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. This newsletter may link to other websites and certain information contained herein has been obtained from third-party sources. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation.

References to any companies, securities, or digital assets are for illustrative purposes only and do not constitute an investment recommendation or offer to provide investment advisory services. Any references to companies are for illustrative purposes only; please see a16z.com/investments. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund which should be read in their entirety.) Past performance is not indicative of future results.

Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Content in this newsletter speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see a16z.com/disclosures for additional important details.

YOU’RE INVITED: Want to host a Tech Week event in NYC or Boston this summer?

The Tech Week team is hosting office hours next week to walk you through how to get listed, promote your event, and make it a great one. RSVP here.

Most founders think fundraising is about selling investors on their vision. I believe it’s not. It’s about making it easy for an investor to say yes.

Here’s some inside baseball: unless you are pitching a solo GP, it’s often not enough for the investor you’re talking to to fall in love with the opportunity. They have to convince the rest of their team to love it too.

This requires writing an internal investment committee memo that translates your story into a case for why the fund should wire you millions of dollars. And they’re doing this for dozens of deals simultaneously, while fielding new pitches, managing portfolio companies, and trying to leave the office before 9pm.

So when it comes down to two promising companies and one evening to write an IC memo, which deal moves forward? The one where the partner has a five-page written document in the founder’s own words? Or the one where they have to reconstruct the narrative from a six-slide deck and some scribbled notes from the week before?

I’ve observed this dynamic play out dozens of times. The decision to bring a company into an investment committee meeting sometimes has less to do with level of conviction and more to do with whether the investor has time to do proper due diligence to write a complete memo. That might sound cynical but it’s reality in a world governed by tradeoffs.

I’ve personally had teams come in where, honestly, I was lukewarm based on the deck. Maybe they were deeply technical and not natural communicators. But when I read their memo, I could see the depth of thinking. The level of rigor changed my read on the entire company.

To put it simply, a great memo helps an investor more accurately price your potential.

And beyond the utility for fundraising, it’s a useful exercise for you, too. The discipline of writing in full prose and structuring your thinking on paper forces a level of clarity that high level bullet points on slides simply don’t require.

This is why every founder I work with at speedrun now writes an investment memo. I think you should write them too.

So what does a great memo look like?

What Goes in the Memo

What follows is a template that I share with every founder I work with.

One-Liner

Your company in a sentence. This should be crisp enough that someone can repeat it from memory to a colleague. In the earliest stages saying something like “We are the X for Y” e.g. “We are the Cursor for Sales” is totally fine. As you get deeper into your journey you’ll want to abandon that framing.

Team Overview

This is the section investors flip to first and your most important one, so you might as well put it in the front. Before they read anything about your market or your product, they want to know who’s building this and why they should care.

Include founder backgrounds with enough detail that an investor can quickly assess relevant experience. What notable things have you accomplished to date? What makes you exceptional? What have you gone 0 to 1 on? If you have a social following, mention it. It can eventually help for distribution. Outline the broader team if applicable, and if you have impressive or relevant advisors, include them.

The core question you’re answering is why is this the right team to win in this market?

Side note: you should hyperlink your Linkedin to your name. The first thing investors want to do when they come across a new opportunity is to pop open their profile to see if they have mutual connections and can therefore reference the team. I know that hyperlinking seems like a silly thing to make sure to do, investors do have opposable thumbs and can type into a search bar, but a quality memo is built by a hundred tiny touches.

Problem and Market Overview

What is the problem, and how big is it? Keep this concrete. Avoid hand-wavy TAM calculations as these are always bullshit. If you can, frame the problem through tangible examples. One memo I got recently did this right by starting with, “sales people at a distributor spend 50% of their time slicing through the messy inbox problem and manually inputting data into a legacy system to generate a quote.” Another option is analogies from spaces with similar root problems (“we’re Tennr for Distributors”) an investor can relate to. This is much better than dubious TAM figures. The best problem statements make the reader think “of course someone should solve this.”

Product Overview

What does the product actually do, and what is your clear wedge into the market you’re attacking? This section is also where the “why now” belongs. What has changed, whether in technology, regulation, behavior, or market structure, that makes this the moment to build a hundred-billion dollar company? Any good idea has likely been tried before, so identifying the current tailwinds is essential.

GTM Strategy

Who is your ideal customer, and how are you getting in front of them? If you have any unique distribution advantages, unfair access to customers, community-driven growth, partnerships, whatever it is, this is where you make that case. Investors want to know you’ve thought about how this thing actually gets sold, not just built.

Business Model

How will you make money? If it’s going to evolve over time as you move upmarket or add product tiers, describe that evolution. Keep it simple but be specific.

Even if you’re early and haven’t fully validated your pricing model with customers yet, demonstrating that you have a working theory here substantiates that you’re building a venture scaled business and not just a cool product.

Traction

This section will vary enormously depending on your stage. At pre-seed, you may only have design partners or letters of intent. At seed, you might have real revenue. At Series A, you should have retention and expansion data.

A few things to include if you have them: an overview of design partners or marquee logos, what they’re paying now and what they could be worth at scale, and whether contracts represent actual billed ARR or contractual commitments not yet billed.

A pet peeve of mine, that’s worth flagging: Do not refer to non-recurring revenue as Annual Recurring Revenue. If your revenue is predictable but not contractually recurring, call it Annualized Run-Rate Revenue and spell that out. If investors think you’re being deceptive with this label, it will break trust.

If you have the data, include monthly revenue charts and monthly customer counts to show acquisition growth and momentum (note: if you don’t show a chart, investors may assume that the growth isn’t great). The other important charts to include are logo retention by cohort and net dollar retention. The a16z blog has excellent breakdowns of the key metrics for both B2B and consumer/prosumer businesses. Include those metrics. Investors know them, and showing you know them too signals sophistication.

If you’re a Series A/B company you may be able to drop in a chart that is going up and to the right to signal that “it’s working” and that you have PMF. Before then, traction can be more about signals (“we’re in a late stage paid pilot with Enterprise Customer A with the opportunity to expand across the entire org”) than raw numbers. Being able to properly contextualize your early traction becomes one of the most important ways that memos trump decks. Use that to your advantage.

Competitive Landscape

Who else is in this market, and why are you going to win? What’s your unique advantage? Investors don’t expect you to have zero competition because great markets attract many great founders. They expect you to have done your research on the competition and have a credible theory for why you’re differentiated.

Vision

What’s the five-year picture? How does the market evolve, and how does your company evolve with it?

Once you’ve penetrated the market with your thin wedge, what is your plan to expand the product surface area and dominate the category? This section is your line of defense when an investor believes your initial product has a limited TAM or is more of a feature than a full product. You should identify how owning your wedge gives you the right to win the category. So if you were building software that will eventually replace ERP (Enterprise Resource Planning) systems you could defend it with something like, “by building the ‘system of action’ with agents above an ERP we can eventually build and replace the ERP.”

This is your chance to show the investor the larger arc of your vision: the insight that turns a useful product into a large and important company.

Next 12 Month Plan and Use of Proceeds

Lay out what you plan to accomplish with the capital you’re raising. Where will it go? What milestones will it unlock?

I would offer one piece of counterintuitive advice: do not put the specific amount you plan to raise in the memo. That should be a live discussion. Writing it down anchors the conversation in a way that can work against you.

When a Memo Matters Most (and When It Doesn’t)

Not every company benefits equally from a memo. I believe the highest-leverage scenarios are:

Contrarian or obscure markets. If you’re building vertical AI for lumber distributors or rail operators or food distribution, the investor probably doesn’t already know your market. They need to be educated on why this is bigger or more interesting than it looks. A memo lets you do that in a way slides can’t.

Technical teams who aren’t natural communicators. Some of the best founders I work with are deeply technical. In a quick intro call, they may not shine. But when they can lay out their thinking in written form, you can see the rigor. The memo becomes the great equalizer between business and technical founders.

Post-meeting follow-ups where the investor is on the fence. If someone is deciding whether to take a second meeting or bring you into their process, having a written document that answers the obvious follow-up questions can tip the balance.

Where a memo matters less: if you’re building in a well-known, fast-moving market like AI Video editing, every informed investor already knows the space. A five-page market breakdown is going to be out of date in three months. In those cases, the memo should lean heavily on team and traction, not market education. Don’t write pages about a market the investor already understands.

The Deck and the Memo Are Not Mutually Exclusive

This isn’t an either/or. You should have both, and they serve different purposes at different moments in the fundraising process.

The teaser deck is your ticket to the meeting. It gets shared before the first call. Its job is to answer two questions for the investor at a thumbnail level: Is this an interesting team? And is this a market I care about? That’s it. It can be five or six slides and skimmable. It should not try to do more than this.

The memo is what you share after the initial meeting. Its job is to capture everything the investor might have missed, provide depth on the market and the strategy, serve as an FAQ, and demonstrate the quality of your thinking. The deck is the teaser. The memo is the substance. And now that tools like Gamma exist, you can actually write the memo first and generate the teaser deck from it. Write the substance, then extract the thumbnail. That’s a better workflow than the other way around.

The goal isn’t to be exhaustive with your materials. It’s to answer the critical questions and get to the next meeting.

A Few More Tactical Notes

On length: One important caution. Particularly at the pre-seed stage, don’t say so much that you give the investor enough rope to hang you. The goal is to answer the critical questions and get to the next meeting. At pre-seed, aim for three pages and make it mostly about team and market opportunity. At seed, five is a good ceiling. Even at Series A, anything beyond eight or ten pages is probably too long. If it feels too long, it probably is. Say just enough. Not more.

When your memo is too long investors may throw it into ChatGPT and ask for a summary. I’ve done this. Many investors I know have done this. So you might as well write something tight enough that it doesn’t need to be compressed by a language model to be useful.

More excitingly, AI is to change what an investment memo can be: One founder recently sent me a long FAQ document along with a system prompt and told me to put both into Claude. The system prompt would generate a memo from the FAQ, and then the investor could ask follow-up questions based on all the source material. I’m not sure I’d recommend this as your primary approach, but it was creative, and it points to a real truth: investors can use AI to process your materials whether you design for that or not.

On writing the memo even if you don’t “need” one: Even if you are one of the select few founders who can raise without written materials, the act of writing the memo is still valuable. It forces you to structure your thinking, identify the weak points in your narrative, and articulate things you may have only ever talked about out loud. Many founders tell me the writing process changed how they thought about their own company. Further, even if you can raise on just your name, if your investors do not deeply understand your vision and direction, it can be harder for them to actually help you build.

The Case for Substance

Long-form is back. Investors want substance. The best founders I work with understand that a well-written document in their own voice is one of the highest-leverage things they can produce during a fundraise. It can be the difference between an investor who’s interested and an investor who falls in love.

Write the memo. Make it easy for people to say yes.

Have questions about investment memos? Reply and let us know. And for more weekly dives into the world of early stage startups, subscribe below.

You are receiving this newsletter since you opted in earlier; if you would like to opt out of future newsletters, you can unsubscribe immediately.

This newsletter is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. This newsletter may link to other websites and certain information contained herein has been obtained from third-party sources. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation.

References to any companies, securities, or digital assets are for illustrative purposes only and do not constitute an investment recommendation or offer to provide investment advisory services. Any references to companies are for illustrative purposes only; please see a16z.com/investments. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund which should be read in their entirety.) Past performance is not indicative of future results.

Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Content in this newsletter speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see a16z.com/disclosures for additional important details.

Tomorrow is the final day to apply for the first-ever cohort of a16z speedrun Alpha, a selective fellowship for exceptional students and recent-grad engineers who want to build startups, either by joining a breakout portfolio company at the ground floor or starting one themselves.

Applicants on Alpha’s Founder Track get a $20k grant and access to up to $250k in follow-up capital, plus priority consideration for a16z speedrun. For those who aren’t yet ready to found, we’ll place you with one of our portfolio companies.

Getting in can change the slope of your career. There’s no reason to wait, so apply today.

In June, the first cohort of Alpha will kick off with a retreat alongside the speedrun team, followed by 8 weeks in person with some of the smartest engineers and operators in the ecosystem. Companies started by Alpha fellows will receive fast-tracked consideration for a16z speedrun.

For those with questions about the program, check out the below Q&A with program co-leads Katia Ameri and Jordan Mazer below.

Q&A

A16Z SPEEDRUN: What led to the creation of Alpha? What need were you seeing?

KATIA AMERI: AI has made it much easier for someone early in their career to actually ship meaningful projects. You don’t need a huge team or a ton of capital to start experimenting. So we were seeing talented people all over building interesting products, sometimes from school, or sometimes from side projects. We realized we wanted a way to meet those people earlier and give them a direct path into startups.

There are two tracks within the fellowship. One is for people who want to join a fast-growing portfolio company and learn by building alongside a founding team. The other is for people who know they want to start something but are still at the earliest idea stage. For them, we give them the capital and infrastructure they need to actually start building.

JORDAN MAZER: There’s a war raging in every channel: poasters on X talking about the next single employee startup that will reach a $1B valuation, Dario going on every broadcast available to postulate about the total collapse of the white collar new grad job market, Citadel recently releasing a report showing that new company creation has gone into hyperdrive.

Common across nearly all of the mediums and personalities is an earnest belief that tectonic upheaval is happening to us in real time. They all agree tools are more capable than ever before, and builders are accumulating leverage at an unprecedented rate. What they seem to disagree on most is when layoffs will occur, and how big the layoffs will be.

Here’s what I see every single day: more and more exceptional students and recent grads struggling to find work they’re excited about. The question is: what happens when exceptional students matriculate into a market that isn’t sure what to do with them? And, more specifically, what happens if those exceptional people happen to be J.A.R.V.I.S.-infused-superhumans who are in no way laden with a) all of the debt of being a big existing company, or b) all the habits of building software in a BC (“before Claude”) world?

Our belief is that many of those young matriculants (or cracked dropouts) are going to do incredible things. Build incredible things. Change the world and markets. And, to be perfectly clear, we’re not just betting with our words.

You’ve announced that the program includes a kickoff retreat, dinners, and AMAs, but what does a random Tuesday in week four actually look like for a Talent Track fellow versus a Founder Track fellow?

KATIA AMERI: For someone on the Talent Track, a random day is mostly their full-time job. They’re embedded at their company, working alongside the team, and the fellowship programming happens around that, usually in the evenings (like cohort dinners or AMAs).

For someone on the Founder Track, the day is focused on building their own idea. We’ll have a shared workspace where fellows can work alongside each other, and there will be structured touchpoints like office hours and cohort dinners. For both tracks, we’re also planning behind-the-scenes visits to portfolio companies at different stages so folks can see how startups operate up close.

The Alpha website says “technical background and experience shipping real work.” But between two candidates with similar resumes, what actually tips the scale? Are you pattern-matching to something specific, or is it more of a “you know it when you see it” thing?

JORDAN MAZER: The strongest candidates will be honest, both with themselves and with us, about what they are trying to achieve. Being honest is harder than it sounds. Because having an honest perspective about what you want requires introspection (very rare) while also understanding the options laid out in front of you (hard for someone at this stage of life!).

Most candidates try to be everything for everyone. They say they’re open to any stage. They say they’re open to any industry. They say they’re open to any problem. Sometimes it’s because they honestly don’t have an informed preference (which doesn’t mean they are open to anything, it just means they don’t know what they want). Often it’s because they aren’t sure what the “best answer” is, irrespective of their actual desires, so they are trying to say the thing that will “keep them in the running.” By trying to be everything to everyone, candidates become totally undifferentiated. It turns otherwise promising applications into forgettable fodder for the rejections folder.

The most interesting prospects will have clarity about their preferences. They will be opinionated (even if they’re wrong!). They’ll tell me they love [X] and that they absolutely abhor [Y], without fearing the risk of affecting my opinion of their fitness for the program. They are honest because, to them, the program is a means to an end: the end of working on what they want to work on. For them, the program itself is only important insofar as it supports their goals and intent.

Some strong candidates will want to work on rockets. Others on high-frequency trading tech. A handful will find the idea of working on AI-driven weapons to be morally objectionable, while some other prospect will tell me it’s a moral imperative. Many will want to found and others will want to join. Common among all of the “strong prospects” we select will be the core capacity to be honest with themselves and with others, and to have conviction in their path.

KATIA AMERI: I’d say what really tips the scale is initiative. We look closely at what someone has actually built. Did you go deep on something because you were curious, and when no one asked you to? We care a lot about demonstrated intellectual curiosity and a bias toward action.

How does the disbursement work between the initial $20K and the rest?

KATIA AMERI: On the Founder Track, we’re often backing people at the individual level, before there’s a company formed. It starts with a $20K grant that we give directly to the fellow upfront. Once they’ve formed an entity, refined the idea, and there’s something more concrete, they become eligible for up to $250K in investment.

Jordan, you’ve been interviewing hundreds of Alpha candidates already. What’s standing out to you about the people who’ve applied so far?

JORDAN MAZER: The range. One of the first founder candidates I spoke with has several planets named after him (from the work he did identifying said planets at NASA… while in high school). Another built a non profit that donated shoes to tens of thousands of impoverished people. She was 13 when she started building it. One was the youngest to ever intern at Lockheed. Another was a challenger-ranked Yasuo main (NA, tho 😭).

Then I get some who ask for extensions on challenges, or who didn’t even read the application and are surprised the program isn’t for internship opportunities or who “just want to found for the summer.”

What about co-founder matching? The Alpha site says you can apply pre-team. In practice, do you expect most Founder Track fellows to find co-founders within the cohort?

KATIA AMERI: For the Founder Track, some people will apply solo, some people will already know who they want to build with. We don’t require a pre-formed team. That said, we’re intentionally creating an environment where the fellows will spend a lot of time together, so we expect that some co-founder relationships will form organically as a result.

What does it mean when we say Founder Track fellows get “priority consideration” for speedrun?

KATIA AMERI: The goal of Alpha is to set you up to be in a strong position for speedrun, which is our flagship program. During the cohort, you’ll have touchpoints with the speedrun team and a clear sense of what we’re looking for. So if, by the end of Alpha, you’ve refined your idea and want to pursue speedrun, you’ll get a guaranteed final-round interview.

Have questions about the Alpha Fellowship? Reply and let us know. And for more weekly dives into the world of early stage startups, subscribe below.

You are receiving this newsletter since you opted in earlier; if you would like to opt out of future newsletters, you can unsubscribe immediately.

This newsletter is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. This newsletter may link to other websites and certain information contained herein has been obtained from third-party sources. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation.

References to any companies, securities, or digital assets are for illustrative purposes only and do not constitute an investment recommendation or offer to provide investment advisory services. Any references to companies are for illustrative purposes only; please see a16z.com/investments. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund which should be read in their entirety.) Past performance is not indicative of future results.

Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Content in this newsletter speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see a16z.com/disclosures for additional important details.

The a16z speedrun team is hiring! We’d love to get your application for:

1. Events & Community Marketing - Run the founder experience for teams accepted into speedrun

2. Product Growth - Help our founders gain traction through modern product-led growth strategies

3. Demand Generation - Help build our investor community

4. Partnerships Manager, Tech Week - Help grow and support our Tech Week partner network

5. Marketing Partner, Alpha - Help us build and grow our exciting new fellowship program for early-career technical talent

You can also see all our openings here. We can’t wait to receive your application.

Now, onto this week’s main story.

The old founder-led sales playbook went something like this:

You manually send DMs on LinkedIn for two hours, hand-write cold emails to everyone you could think of, track everything in a spreadsheet, pray. If you were selling to enterprises, all you had to do was multiply your expected pain by twelve and add a procurement department.

That playbook is (thankfully) kaput. AI-native founders are now selling to banks, hospitals, law firms, and universities with zero dedicated sales hires. The type of sales processes that used to take 18 months can now go much, much faster.

The unlock comes from two capabilities converging:

1. computer use agents are finally good enough that they can work autonomously

2. automated outbound pipelines are cheap and easily accessible.

In 12 months our inboxes are likely to be spam-filled and stuffed with AI slop. For now though, most of the world hasn’t figured out how to do this, so there is a brief, glorious moment where founder-led sales can use AI to accelerate revenue growth.

Here’s how speedrun founders are building the stack.

Your AI SDR Lives in a Browser Tab

Computer use is the capability that changes everything about solo founder sales. These are AI agents that can see your screen, click buttons, fill out forms, and navigate web apps. By freeing yourself from the constraints of APIs and MCP, you can build a janky but serviceable SDR with a minimal amount of effort.

Tools like Claude Cowork, OpenAI Operator, and similar browser-control agents can do most of the things you can do in Chrome. And most prospecting workflows live in browser-based tools like LinkedIn Sales Navigator and CRMs that either don’t have APIs or charge enterprise prices for access. Computer use bypasses all of that.

Shawn Tsao, cofounder and CEO of Snapp (SR006), runs Cowork on a browser tab pointed at Sales Navigator. It reads his ICP filters, clicks through profiles, and sends connection requests on autopilot, maxing out LinkedIn’s 200-per-week limit. Each lead gets added to a Sales Nav list spreadsheet, which then flows into Dripify for automated LinkedIn drip sequences. His own review: “A bit scuffed, but it saves a ton of time.”

How to replicate this today:

Define your ICP in Sales Navigator using title, company size, industry, and geography filters. Open Cowork (or your browser agent of choice) on that Sales Nav tab. Instruct it in plain English to connect with matching profiles and add them to a named lead list. Then pipe that list into a LinkedIn automation tool like Dripify or Expandi for follow-up sequences. Treat the browser agent as a junior SDR who can do anything you can do in a browser. You just tell it what to do.

To be clear: this will almost certainly break or have weird edge cases. But that’s ok! If you’re an early stage founder, perfect processes are reserved for those with product-market fit.

Tsao adds one more tip:

“I also ended up building my own AI Agent that doesn’t require Sales Navigator but uses publicly found information to get cold leads and even writes out the exact email and places it into my Gmail drafts.”

From Lead to Demo, No Humans Required

The second layer moves beyond simple prospecting into full pipeline automation. AI manages sourcing, enrichment, outreach, and response routing end to end. An orchestration stack typically combines tools like Clay, Lemlist, and Attio as the central hub, connecting data sources, AI enrichment, and outbound channels into a single workflow.

Silvia Chen, cofounder and CEO of Bilrost (SR006), built this exact type of system to sell loan processing automation software to commercial and business-purpose real estate lenders. Leads are sourced, enriched and qualified in Clay, synced into Lemlist for personalized outreach, and routed through an AI-driven decision tree. Each response triggers a different path: message viewed, reply received, no response, objection raised. Each generates its own tailored follow-up sequence, and the system runs autonomously until a prospect is ready for a live conversation.

Chen told us that now her “LinkedIn DMs are so flooded I can’t open the app.” When she launched a last-minute, conference-specific campaign ahead of a major industry event, her calendar filled up from landing to takeoff with high-quality, in-person meetings—selling into banks, one of the most conservative and difficult buyer segments.

How to replicate this today:

Start by building your lead list in a tool like Clay or 11x, which can pull from LinkedIn, Crunchbase, and company databases based on your ICP filters. Use Clay’s AI enrichment to research each lead, write personalized first lines, and identify the right entry point at each organization. Route enriched leads into outreach tools like Instantly or Smartlead for email, or Dripify for LinkedIn, with conditional sequences. Set up branching logic so different follow-ups trigger based on opens, replies, and objections. Then insert yourself only when a lead is qualified and ready for a live conversation. (And if this all sounds too complicated, just copy and paste this paragraph into a coding chatbot of your choice, and you can build it in an afternoon.)

The goal is that you should only be doing the part of sales that requires you personally, the demo, the relationship, and the close. Everything upstream of that can run mostly autonomously.

The Trust Layer: What Automation Can Accelerate But Not Replace

There’s a piece of enterprise sales that no automation stack can skip entirely: Every industry has what GalaxySail cofounder Ruming Zhen (SR006) calls “tokens of trust.”

These are specific artifacts buyers need to see before they’ll engage seriously. Hospitals need published research and conference posters. Schools need pilot results. Banks need compliance documentation. Every industry will have some signifiers of security or quality that it is important to have

AI can’t eliminate the need for these artifacts, but it can compress the timeline for producing them from months to days. Multiple speedrun founders described generating SOC 2 readiness documentation and security architecture summaries using AI rather than hiring consultants. Several discovered that simply getting a letter of intent from a provider like Vanta, showing you’re in the process of obtaining certification, was enough to unblock deals. One current speedrun founder got the letter the same morning he reached out, and it moved a stalled enterprise conversation forward immediately.

The bottleneck in enterprise sales isn’t usually the pitch. It is getting the customer to believe in you comically early in your journey. These tokens of trust give buyers permission to say yes. AI compresses that production cycle dramatically. Identify the one or two trust artifacts your buyer needs, then use AI to produce them in days. Think of these as specialized forms of sales enablement!

Other AI GTM Tools SR Founders Are Using

We asked a wide group of founders from SR006 what AI tools and workflows they’re using for GTM and sales.

Here are five more interesting approaches:

“We built a custom GTM dashboard to enrich public data with founders’ LinkedIn data. This is important since the legal industry operates on warm connections.”
—Ke Ma, cofounder and COO of Concorda (SR006)

“We run a multi-model, agentic GTM stack. GPT-4.1 handles structured reasoning and tool use, Claude 3.7 synthesizes long-form thesis shifts, and Gemini 1.5 processes large context windows for portfolio and filing analysis. Tavily + Apify feed real-time capital signals, a lightweight LangGraph agent loop continuously reprioritizes deployment probability, and outbound + pre-call briefs are regenerated dynamically. CRM is just the execution layer. Inference drives pipeline.”
—Mohamed Mohamed, founder and CEO of Smart Bricks (SR006)

“We’ve found that cold-email outbound is becoming an increasingly tough channel (though Instantly is great) and that trying out many different AI tools can be more distracting than helpful. Going deep with a couple tools keeps us focused. Our ‘outward-facing’ stack for lead-sourcing, enrichment, and outbound centers is Exa + Superhuman. Our ‘inward-facing’ stack for note-taking/ops, CRM, and PM is Notion + Slack + OpenAI suite.”
—Andy Baran, cofounder and CEO of a stealth company in SR006

“At Quanto, we’ve put a lot of our founder-led sales meetings notes and our internal team meeting notes into a custom GPT project. The folks on our team that do sales / customer success will query that first with any inbounds from current and prospective customers to help make it sound like me as the “first line of reasoning.” It has been helpful in scaling and trusting the team to interact directly with clients.”
—Anderson Petergeorge, cofounder and business lead of Quanto (SR006)

“We manage a 5-agent AI sales org. They run async: plugged into our sales stack, sourcing leads, researching prospects, qualifying fit, and writing custom outreach for each. We steer the whole team over iMessage.”
—Stefano Delmanto, cofounder of Mercury (SR006)

The minimum viable sales team

Founder-led sales used to mean the founder does everything manually. Now it means the founder orchestrates agents that do everything. Founders in speedrun are now running sophisticated GTM outreach programs with 1/10 of the resources required five years ago.

The new stack is a browser agent for prospecting, AI tooling like 11x or Clay for enrichment, automated sequencing for outreach just using an email agent, and AI for trust artifacts.

In the earliest stage, no one’s title is “sales,” so why not offload more of that work to AI?

Have questions about AI-powered sales and GTM? Reply and let us know. And for more weekly dives into the world of early stage startups, subscribe below.

You are receiving this newsletter since you opted in earlier; if you would like to opt out of future newsletters, you can unsubscribe immediately.

This newsletter is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. This newsletter may link to other websites and certain information contained herein has been obtained from third-party sources. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation.

References to any companies, securities, or digital assets are for illustrative purposes only and do not constitute an investment recommendation or offer to provide investment advisory services. Any references to companies are for illustrative purposes only; please see a16z.com/investments. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund which should be read in their entirety.) Past performance is not indicative of future results.

Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Content in this newsletter speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see a16z.com/disclosures for additional important details.

Earlier this month, OpenAI’s VP of Science Kevin Weil sat down for a fireside chat before an audience of a16z speedrun founders, hosted by SR visiting partner Sam Shank. After leading product teams at Twitter, Facebook, and Instagram, Kevin is now focused on accelerating scientific discovery with AI.

The conversation ranged from how AI capabilities are progressing to the future of robotic labs, how founders should think about building with AI, and why this might be the best moment in history to start a company.

Watch the talk in full below, or read on for five moments that stood out to us.

1) On the rapid nature of AI capability increases

Weil opened the conversation by pointing out that AI models have now solved open mathematics problems that humans had never solved before. When Sam Shank asked him to unpack how AI capabilities are progressing, Weil offered a framework that applies far beyond science:

“If there’s one thing I’ve learned over the last few years, it’s that you go very quickly from ‘models could never do this thing, it is beyond the capability of AI today’ to ‘models can just barely do this thing, and it kind of sucks at it, and it’s wrong most of the time, but you get these glimmers.’ Maybe it only works 5 or 10% of the time. And then six to twelve months later, it’s like ‘models are great at this thing, and I would always use AI anytime I ever do that again.’ In eval language, you go very quickly from zero to 5 or 10% to 60 or 80%.”

The implication for founders: if you see a capability that barely works today, it’s probably six to twelve months away from working well. Weil says we’re in exactly that middle phase right now with frontier science.

2) The science of 2050, but in 2030

Weil’s biggest bet at OpenAI is that AI will accelerate scientific discovery in ways we can physically feel: new materials, personalized medicine, fusion power. He laid out a specific vision for how that happens:

“The science of the future will definitely involve robotic labs and reinforcement learning loops that go through the real world. The model is thinking, maybe running a simulation, thinking some more, refining the experiment it can run using the best possible parameters, and then sending that to a bunch of robotic labs, which you can scale horizontally. The experiments run in real life. The results come back to the model. The model thinks, runs more simulations, thinks. You have multiple loops: tight loops with the model thinking and the simulation, and longer loops that go through the real world.”

“You have robotic labs that can scale horizontally, that can run 24 hours a day. They’re not grad students pipetting things who need to take breaks and sleep. And then the grad students can do things that are much more leveraging of what makes us human than pipetting things.”

Weil says this isn’t speculative. Robotic labs already exist, open math problems are already being solved by AI, and the pace of progress makes him confident the full loop won’t take long to close.

3) “I just wasted an hour”

When Shank asked about the moment Weil realized the world had fundamentally shifted, he didn’t point to a weekend project. He pointed to a meeting:

“I was sitting in a meeting and I closed my laptop because we’re all trying to be better about not multitasking. I had not gotten a Codex job running before I closed my laptop, and I was like, ‘Shit, I just wasted an hour.’ Not because of the meeting, but because I could have been multitasking during that hour. My Codex agent could have been fixing a bug or implementing a feature or doing something for me.”

“The same thing before you go to bed at night: ‘Okay, what really hard task can I give Codex and just let it chunk away for ten hours?’ If you’re really good at it, you’re not just juggling one job. You’ve got three or four things running in parallel across different work trees.”

This is a concrete picture of how the highest-performing people at OpenAI are already working: always running background tasks, always parallelizing, treating dead time as lost compute.

4) “The most fertile ground for startups that has ever been”

When the conversation turned to why Weil’s bullish on the current startup environment, Weil made the case that the sheer novelty of AI capabilities creates an unprecedented window for founders:

“Today, models can do something that computers have never been able to do in the history of computers. And in another month, that’s going to happen again. And in another month, it’s going to happen again. We don’t always know what’s coming. Sometimes these things are emergent capabilities of models as we build them. And when you’re surprised, everyone in the world that uses a model has this new capability that nobody has ever had before.”

“One of the reasons I’m so bullish about startups right now: there are so many new capabilities and the world doesn’t quite know what’s possible. OpenAI doesn’t always know what’s possible. We’re not going to have all the ideas. It is just the most fertile ground for startups that there has ever been.”

Weil connects this to a practical point: this moment selects for people who are high agency. If you have an interesting idea, you have no excuse not to get Codex working on it in the background while you do whatever you were going to do that morning. You might have a working prototype by end of day.

5) On mistakes founders make when building using AI

Shank asked what mistakes he sees founders make when building on top of AI. Weil’s answer was specific and practical.

“A lot of things today turn out best when you use an ensemble of models. You might have an initial model that’s orchestrating, putting a plan together, understanding what you should do to answer the question. Then you have different models, maybe cheaper models that are trained to do one thing really well. The orchestration model is calling the other models. I don’t see people doing that enough.”

“Behind the scenes, we use ensembles of models in lots of places. Trying to use small models where you can, bigger models where you need to, and then have them all work together versus just prompt-engineering one giant prompt and hoping the answer is right.”

For now, the gap between one-shotting a complex flow and reliably handling it with an ensemble is significant, and he sees too many startups leaving reliability on the table.

Thanks to OpenAI’s Kevin Weil for speaking with our founders. And for more weekly dives into the world of early stage startups, subscribe below.

You are receiving this newsletter since you opted in earlier; if you would like to opt out of future newsletters, you can unsubscribe immediately.

This newsletter is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. This newsletter may link to other websites and certain information contained herein has been obtained from third-party sources. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation.

References to any companies, securities, or digital assets are for illustrative purposes only and do not constitute an investment recommendation or offer to provide investment advisory services. Any references to companies are for illustrative purposes only; please see a16z.com/investments. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund which should be read in their entirety.) Past performance is not indicative of future results.

Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Content in this newsletter speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see a16z.com/disclosures for additional important details.

🙅 Jon Lai argues that “learning to say no” to customers is critical for early teams

🔎 Last week, readers loved Limy co-founder Aviv Shamny’s tips for standing out in AI-powered searches. The Limy team is hosting a webinar on March 10th to share more knowledge about search in the Agent Economy. Join to stay ahead of the curve.

🎡 The next AI Faire will take place March 6th in San Francisco. Sign up if you want to see ~25 teams building the future of AI B2B tools.

💼 Join our talent network for more opportunities. If we see a fit for you, we’ll intro you to relevant founders in the portfolio

This week’s post is a guest essay from a16z speedrun visiting partner Sam Shank, who co-founded and led HotelTonight, which sold to AirBnB in 2019. Sam makes a case for applying extreme focus to the metrics they track and prioritize.

I’ve started three companies. The first two didn’t find product market fit. The third one created a new category in travel and sold for over $400 million. The biggest difference between them had nothing to do with the market, the team, or the product. It had to do with a simple spreadsheet.

My second company was DealBase, a travel deals search engine. It was a complex business so I built what felt natural: A weekly dashboard to track our progress. This sounds like a reasonable thing to do! Except we just kept adding to it. Every week, somebody would say “we should also be tracking X” and we’d add another column. Eventually, we were staring at roughly 80 metrics every single week. Perhaps you too have had a similar explosion in stuff to track, a phenomena I think of as metrics slop. Staring at that spreadsheet, I had no idea if the business was actually doing well. We were measuring everything and understanding nothing.

At HotelTonight, we went all in on one number: total transactions. We just wanted to increase the number of transactions that we brought to our hotel partners. That’s it.

DealBase didn’t work out while HotelTonight became one of the fastest-growing travel apps in the world. I don’t think that’s a coincidence. And in the years since, as an investor, the pattern has only gotten clearer to me. The best early-stage founders I meet share a kind of painful, almost reckless simplicity about how they run their companies. They’re all in on improving one metric while de-emphasizing everything else.

This idea sounds obvious. And that’s exactly why most founders fail to do it.

How Simplification Saved HotelTonight

It wasn’t like I just hopped into weekly all-hands and said “increase transaction volume, team” and called it a day. Instead, I became obsessed with simplification across every dimension of the business. The point of the metric is to be a forcing function, the lens by which all company choices are examined. For example, we engineered the product to maximize volume. It took eight seconds to book a hotel, three taps.

The early days tested that conviction.

When we launched, people were interested in what we were building. They downloaded the app. They just weren’t booking. We had targeted 100 transactions in our first month, and we got 55. About half of those were us booking hotels for our friends and family, or essentially forcing them to use the product.

That was a gut-check moment. Do we even have a business here? We had set a clear goal, missed it, and now had to decide what to do. So, we did as all great tech startups do, we doubled down. It worked. It just took a lot more effort than we expected.

For years after that, transaction volume was the only number that mattered (though we shifted to revenue, and then profitability as the company matured.) But in the early days, when we needed to prove to our hotel partners that the flywheel was spinning, the simplicity of that one metric kept the entire team focused on what actually mattered.

This approach isn’t unique to HotelTonight. But for other businesses, and other founders, how should you think about finding that “only number that matters?”

The Pattern Across Iconic Companies

Many of the most successful startups of the last two decades share a version of the same discipline, and the metric they chose varied by business type. But the principle was consistent: find the one number that, if it goes up, the business is absolutely going to be in a position to be sustainable.

B2B SaaS companies tend to rally around ARR (but that may be wrong).

You can collect design partners, letters of intent, or verbal commitments. But if nobody is paying, it doesn’t count, so I understand why people focus on ARR so much. That said, when you are building productivity software and are trying to enable a new way of working, then ARR is a trailing indicator. Instead, you should focus on usage.

Slack is the perfect example of this. They discovered that once a team exchanged 2,000 messages, 93% of them continued to use Slack. That number became the focus of every product decision, onboarding tweak, and engineering sprint. The result was among the fastest any B2B SaaS company had ever reached a $1 billion valuation at the time. Every product has some point of critical mass, where usage crosses the productivity rubicon, and suddenly your customers can’t imagine going back to the “old way” of doing things.

Marketplace businesses should obsess over throughput. Getting the flywheel spinning is everything in the early days. Whether that is transaction volume or gross booking value, you should be all in on whatever captures whether supply and demand are connecting.

Airbnb almost died before they found this focus. The founders had maxed-out credit cards and couldn’t get investors to return their calls. Paul Graham’s advice reframed their entire approach: it’s better to have 100 customers that love you than a million customers that sort of like you. They narrowed everything to one metric (nights booked) and one market (New York City). Brian Chesky flew from Mountain View to NYC, went door-to-door photographing listings, and even slept in hosts’ apartments. Their financial goal was “ramen profitability:” $4,000 per month to cover rent and food, a number they taped to their bathroom mirror. Booking fees went from $460 to $897 to $1,428 in three consecutive weeks, and the flywheel caught.

Consumer companies should focus on engagement and retention before monetization. For a lot of social products, DAU/MAU ratios and audience size are what matter because once you have the audience and the retention, you can monetize in a bunch of different ways.

WhatsApp made the most extreme version of the philosophy anyone has ever seen. Jan Koum tracked monthly active users and message reliability. That’s it. In 2013, competitors LINE and KakaoTalk were generating $338 million and $200 million in revenue respectively. WhatsApp earned $10.2 million total. Koum even charged $1 per year for the app as an intentional brake on growth, so his team could focus on building servers that wouldn’t crash and a product that wouldn’t drop messages. When Facebook acquired WhatsApp for $19 billion, users were opening the app an average of 23 times per day and it was handling 50 billion messages daily. That’s more than global SMS volume.

Why People Don’t Do This

If this advice is so obvious, why do most founders ignore it? Because following it requires accepting a kind of suffering that goes against every natural instinct.

Focusing on one metric means watching money leave your account every month and choosing not to panic about it. It means firing someone who’s well-liked but doesn’t move the number. It means hearing public criticism and deciding it doesn’t matter yet.

Venture capital exists to give you permission to make these trade-offs. When a VC writes you a check, what they’re really saying is: you can now make payroll for two years. Go build toward proving one thing. That permission, which is agonizing, is the actual product you’re buying when you take early-stage venture money.

The exercise that sharpens all of this is taking out the red pen. Periodically, look at everything the company is working on and ask: are we working on the very best things right now to grow that number, or should we shift resources?

Why I could be wrong

The strongest counterargument to this approach is that myopic focus leads to bad outcomes. There are many examples in Silicon Valley of companies that were just flashes in the pan.

However, the answer isn’t more metrics. The answer is building a founder’s intuition to what matters beyond the metric.

Good founders understand intuitively where the guardrails are. They know when a customer complaint signals a systemic problem and when it’s noise to brush off. They don’t need a customer satisfaction metric. That would be the accountant approach, and it leads you right back to 50 KPIs and a team that doesn’t know what to focus on.

The CEO can privately monitor those guardrail signals. You may have a private dashboard with those 80 metrics! But you don’t need to focus the whole team on them. The real skill is choosing a number where growth and sustainability aren’t in conflict—and then trusting your judgment on everything else.

The Action Step

If you believe any of this, here’s what to do when you close this email or switch tabs.

Pick your one metric. The specific number depends on your business type but the principle is the same. Choose a metric where, if it goes up, the business is clearly going to survive and grow.

Live and breathe the metric. Make sure every person on the team knows the number. It should be top of mind every day, every moment. You should wake up wanting to see what that number is. You should obsess over it.

Make every initiative ladder up. Every project, every sprint, every hire should connect to growing that metric. Sometimes the connection is direct. Sometimes it’s indirect. But if someone can’t articulate how their work ladders up to the one number, that’s a problem.

Take out the red pen. Regularly audit what the company is working on. Ask: are these the very best things we could be doing to grow the metric? If not, make changes. This is where the pain lives, but this is also where the focus compounds.

Have questions for Sam? Reply and let us know. And for more weekly dives into the world of early stage startups, subscribe below.

You are receiving this newsletter since you opted in earlier; if you would like to opt out of future newsletters, you can unsubscribe immediately.

This newsletter is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. This newsletter may link to other websites and certain information contained herein has been obtained from third-party sources. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation.

References to any companies, securities, or digital assets are for illustrative purposes only and do not constitute an investment recommendation or offer to provide investment advisory services. Any references to companies are for illustrative purposes only; please see a16z.com/investments. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund which should be read in their entirety.) Past performance is not indicative of future results.

Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Content in this newsletter speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see a16z.com/disclosures for additional important details.

We just launched a16z speedrun Alpha, a selective fellowship for exceptional students and recent-grad engineers who want to build startups, either by joining a breakout portfolio company at the ground floor or starting one themselves.

Getting in earlier can change the slope of your career. If you’re already shipping, there’s no reason to wait.

Alpha has two tracks:

1. You can either join a fast-growing portfolio company full-time, and get hired directly as an engineer building core product alongside founders and early teams

2. Or you can receive $20K equity-free grant upfront to go all in on exploring an idea, with eligibility for up to $250K in investment from us later.

In June, the first cohort of Alpha will kick off with a retreat alongside the speedrun team, followed by 8 weeks in person with some of the smartest engineers and operators in the ecosystem. Companies started by Alpha fellows will receive fast-tracked consideration for a16z speedrun.

We’d love to have you apply. And if you know someone who should see this, please feel free to forward.

Become a fellow —> alpha.a16zspeedrun.com

Jack Altman is the managing partner of Alt Capital, a $275 million early-stage fund, and the co-founder and chairman of Lattice, an HR platform used by more than 5,000 companies. He is also the host of the podcast Uncapped with Jack Altman.

In a recent fireside chat before an audience of a16z speedrun founders, Jack Altman sat down with a16z speedrun investor Marcus Segal to discuss what actually matters at the seed stage.

Watch the full talk:

Our top takeaways below 👇

1) Your first ten hires should be diamonds in the rough

When Segal asked about hiring a startup’s first employees, Altman was blunt about the math. A truly great early hire has to be someone the rest of the market hasn’t figured out yet.

“In order to be a great company, you need a high percentage of your first ten employees to be truly great. There are almost no examples of great companies where that didn’t happen. So that’s the task at hand.

“You also have to take the pragmatic view that for an obviously great employee, it doesn’t make that much sense for them to be the sixth employee of a startup. If they are that obviously great, they’re going to get ten gazillion dollars from a frontier lab or SpaceX or whatever, or they’re going to get an exec job at some super hot series D company, or they’re going to be a founder. So what you’re really looking for is diamonds in the rough. People who are in fact great, but are not legibly great to the entire world. It’s the only way.”

Altman saw the same pattern when hiring Lattice’s early sales team. The reps they pulled from large orgs wanted a playbook and a process that didn’t exist yet. The ones who worked were creative and entrepreneurial, willing to figure it out as they went.

2) When stuff works, it works fast

Lattice’s performance review product, the one that made the company, was actually its third pivot. Altman said the first two attempts had the right problem but the wrong solution.

“Most of the time when stuff works, it works fast. And I still believe that. That’s not a hard and fast rule. There are things that take time. There are situations where you don’t quite have product market fit, and it is actually right on the other side of a few more features or a little bit more sanding of the edges. But directionally speaking, when people want what you’re selling, you know quickly.

“We just weren’t getting the pull. It always seemed like one more feature, one more feature. And it was just never there. And then on a third attempt, everything was broken and we sold it anyway.”

Altman shared a vivid example of how scrappy that early traction looked in practice. When Lattice launched its first paying performance review customers, the product didn’t yet have a way to close a review cycle. His co-founder Eric Koslow was programming the ending while the first customers were already running their reviews. “They were on the beginning of the roller coaster that didn’t yet have the end,” Altman says.

3) Have a North Star for your product, and protect it

The tension between listening to customers and protecting your roadmap came up repeatedly. Altman acknowledged there’s no clean rule for it. Both extremes fail.

“It’s obviously wrong to scramble to [respond to] every single customer request. It’s also wrong to ignore it all… like, ‘No, I know best.’ You’ve got to have some North Star vision of your product. But customers are good at helping fill in the gaps within that outline.

Where it gets really hard is when a single deal threatens to pull you off course.

“One of the trickiest things is when some big enterprise comes along like, ‘Hey, we’ll give you $400,000 a year if you do this thing that zero other customers have asked you for,’ and right now you’re at $220,000 of ARR and you’re like, ‘Oh, I could triple if I just derail myself for like a month.’ It’s kind of tempting. And usually that’s a hard call.”

Segal added: “But it’s never a month, and it’s never their last ask.” Altman agreed, but pushed back against a blanket rule. He pointed to Stripe, whose early roadmap was shaped by large customers like Shopify and Lyft, and those features ended up helping everyone else. The judgment call is whether the request opens up a broader market or just serves one account.

4) Do your work proactively, not reactively

When Segal asked for one habit that compounds over time, Altman’s answer was about protecting your attention.

“You want to do as much of your work proactively versus reactively as possible. As time goes on, it is really easy to just work out of your inbox rather than off your to-do list, or take meetings that somebody else asked for rather than ones that you asked for. A lot of people end up doing reactive work ninety percent of the time, and it should be like nine percent of the time.”

Altman connected this to a broader point about staying close to the product. Once a company gets past about twenty people, the CEO’s job shifts from building the product to building the machine that builds the product. That transition makes it easy to lose touch with what customers actually experience. Someone described this to him as “taking ice cores all over the company all the time,” constantly sampling to make sure the micro still works.

5) No advice is right for 100% of companies

Altman offered this useful meta-observation about startup advice in general:

“Any advice that you get that’s given to a group of companies is doing its best to fit to the highest number possible. But there’s no advice that’s right for one hundred percent of companies. If you listen to it all as dogma, something’s going to be wrong for sure.”

He applied this principle to his own answers throughout the talk. On hiring: “This is eighty percent confidence advice. There are tons of other examples.” On building scrappy early product: “Standards from customers are a little bit higher now. So I’m less confident on this.”

On fundraising, he was more direct: follow the template your accelerator gives you, run a compressed timeline, and don’t let investors drag you into a premature process: “If you don’t have a term sheet, you’re not being preempted.”

That’s it for this week. Thanks to Jack for sharing his insights with our founders.

Like this post? Forward it to your team! And for more weekly dives into the world of early stage startups, subscribe below.

You are receiving this newsletter since you opted in earlier; if you would like to opt out of future newsletters, you can unsubscribe immediately.

This newsletter is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. This newsletter may link to other websites and certain information contained herein has been obtained from third-party sources. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation.

References to any companies, securities, or digital assets are for illustrative purposes only and do not constitute an investment recommendation or offer to provide investment advisory services. Any references to companies are for illustrative purposes only; please see a16z.com/investments. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund which should be read in their entirety.) Past performance is not indicative of future results.

Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Content in this newsletter speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see a16z.com/disclosures for additional important details.

🚀 Our team’s Bella Nazzari has launched a new Substack, open to work, that curates new job openings across the tech startup ecosystem. Go sign up!

🎡 Macy Mills announced that the next AI Faire will take place March 6th in San Francisco. Sign up if you want to see ~25 teams building the future of AI B2B tools.

📩 Andrew Chen says he’s still looking for his perfect AI-powered email app

💼 Join our talent network for more opportunities. If we see a fit for you, we’ll intro you to relevant founders in the portfolio

Since the dawn of the online age, one of the biggest marketing challenges has been how to optimize your website to get machines to recommend you to humans. However, how you actually do that is very different then traditional SEO.

The Paradigm Shift

Aviv Shamny, a co-founder of AI search optimization startup Limy (SR005), shared research on 80 million lines of clickstream data found that nearly 90% of sources these AI engines cited were on page 21 or beyond in Google results.This tracks with Ahrefs’ study showing only 12% of URLs cited by AI tools overlap with Google’s top 10 results.

All of which is to say, the game has shifted significantly from two years ago. Your SEO-optimized content ranking on page one could be completely ignored by AI agents.

But with the changes also comes new opportunities, “We see much higher conversion rates from users coming in from AI search engines,” Shamny told us. “Lower volumes, but high conversion rates. You do the whole discovery and consideration process with ChatGPT, then you just go to the website and buy.”

Recent studies back up Shamny’s POV. Adobe reports that by Jan 2026, AI traffic drives 32% revenue per visit better than non-AI traffic—a complete reversal from earlier in the year—while holiday season AI-driven traffic surged 693% year-over-year. Microsoft’s analysis of 1,200+ websites found AI platforms convert at 3-17x the rate of traditional channels, with Copilot leading at 17x, Perplexity at 7x, and Gemini at 3-4x.

So the channel is real and the opportunity is growing. But how do you actually get your startup to appear in AI-powered search results?

How AI Search Actually Works

To optimize for AI, you need to understand how it differs from Google search. When a user asks ChatGPT “What are the best running shoes for marathon training?” the model doesn’t search for that exact phrase. It translates the query into a system prompt—something like “men running shoes reviews 2026”—which can represent 200-400 different user queries.

The AI then searches for vector spaces similar to that system prompt, looking for content that semantically matches the underlying intent. This is why keyword optimization alone fails. AI prioritizes semantic meaning over keywords, citation authority over backlinks, and direct answers over click-optimized headlines.

How to get your results to appear

For SEO, the common advice was to just pump out content. That rarely works anymore. “If you’re just writing fluffy blogs about your perspective around a certain category, that will never do enough,” Shamny explains. “What actually resonates with AI search engines is problem-solving.”

Written content still works, but the structure matters. You should structure your content as:

Question → Direct Answer → Evidence → Follow-up Questions.

Research shows this approach can boost AI visibility by up to 40%. Put your direct answer in the first 40-60 words, include statistics every 150-200 words, and use FAQ formatting with proper schema markup.

“Content distribution is the most important thing here,” Shamny says. “Not just having it in one place, but having it where AI search engines look.”

Where do they look? Zenith’s research found ChatGPT cited Reddit in 81% of technical queries. The nuance is that AI doesn’t favor viral posts! The median age of a cited Reddit post is 1.5 years. AI wants evergreen, definitive answers. One strategy could be creating genuinely helpful responses in niche subreddits relevant to your category.

Still, the underlying search algorithms are changing much more rapidly than SEO does anymore. Each new provider and each new model update can shift the preferred source of truth from Reddit to YouTube to niche technical blogs. So an important part of the aspect is not merely relying on placing content. You also should build out technical infrastructure to measure everything.

Technical Infrastructure

“You have to get comfortable with advertising to machines and not to people,” Shamny told us.

The technical foundation starts with your robots.txt file. Many sites blocked AI crawlers during the training data debates of 2023-24. Check yours and ensure GPTBot, ClaudeBot, and PerplexityBot have access. Since ChatGPT uses Bing for real-time data, submit your sitemap through Bing Webmaster Tools. The emerging standard is llms.txt—a new file format (like robots.txt) that tells AI crawlers which pages to prioritize. It sits in your root directory, formatted in Markdown, listing your highest-value URLs. Finally, ensure server-side rendering. AI crawlers don’t execute JavaScript—analysis shows ~11.5% of ChatGPT’s requests are JS files that go unused. Your critical content needs to be in the HTML source.

When all of this is implemented the results can be dramatic and fast. Tastewise, a food tech company, reported a 600% visibility spike in one week after updating their Gen Z food trends report with 2024 data, restructuring for LLM summarization, and adding FAQ schema markup. The article went from 42 AI impressions to 294, securing citations in ChatGPT and Gemini.

It can take surprisingly little work to start ranking higher, which is why this is such a compelling growth meta right now.

The Temporary Advantage

It’s still early days. Most companies are applying 2015 SEO playbooks to 2025 AI search which means there is an opportunity for early-stage startups to build their AI search presence with ease compared to the entrenched competition in SEO.

The window also exists because AI platforms develop preferences for consistent, authoritative sources. The startups that figure this out now are establishing citation patterns that become increasingly difficult for competitors to displace, creating a potentially rare, compounding growth asset.

As for measurement, you should track AI agent traffic to your site, not just user traffic. For robust tracking you’ll want specialized software, but start off by doing simple manual tests—ask ChatGPT questions in your category and see who gets recommended.

This won’t stay easy forever. Eventually, every growth channel gets competitive and expensive. But right now, while most companies are still optimizing for page one of Google that nobody reads anymore, there’s actual arbitrage potential available here.

Have questions about GEO? Reply and let us know. And for more weekly dives into the world of early stage startups, subscribe below.

You are receiving this newsletter since you opted in earlier; if you would like to opt out of future newsletters, you can unsubscribe immediately.

This newsletter is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. This newsletter may link to other websites and certain information contained herein has been obtained from third-party sources. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation.

References to any companies, securities, or digital assets are for illustrative purposes only and do not constitute an investment recommendation or offer to provide investment advisory services. Any references to companies are for illustrative purposes only; please see a16z.com/investments. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund which should be read in their entirety.) Past performance is not indicative of future results.

Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Content in this newsletter speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see a16z.com/disclosures for additional important details.

Chris Best is the CEO and cofounder of Substack, a platform you’re likely aware of.

In a fireside chat with a16z speedrun GP Andrew Chen, Chris shares his founding story, discusses how to navigate controversy, and makes the case that having strong values matters more than pleasing everyone.

Watch the full talk:

Our top takeaways below 👇

1) Look for missionaries, not mercenaries

Best traces Substack’s origins to an essay. He was writing about his frustrations with the media economy on the internet. As he put it, he was “going on this unhinged tirade about how Twitter was driving us all crazy.” He sent it to his friend and future cofounder, Hamish McKenzie.

“Hamish kind of let me down gently. He said, ‘Your essay sucks. It’s boring. Anybody can complain. Everybody knows the newspaper industry is dying. The more interesting question is: what would you do about it? How could it be different?’ For me, the idea to start a company came from the idea to do something that mattered.”

Best invoked Jeff Bezos’s framing of mercenaries versus missionaries. The missionaries tend to win, he argued, because they’re the ones who stick around when it gets hard.

“One of the counterintuitive things is that the people who are in it for the love of the game often end up in the ranks of the very most successful. Along the way there are moments: ‘Do you want to sell? Do you want to get out? Do you want to make a buck?’ Those things come. I don’t think Substack would have worked if we were just trying to make money. It’s not the company you would start if you’re looking for the fastest way to riches.”

He also pointed out a hiring advantage that early-stage founders tend to overlook: when your company isn’t cool yet, the only people who show up are true believers:

“The only reason a brilliant software person was going to come work at Substack was because they believed in what we were doing, and that’s an advantage you should treasure while you have it.”

2) Have a sci-fi vision and a short-term plan

When Andrew Chen asked about Substack’s product evolution, Best laid out his framework for how to build a company. It produced one of the most quotable lines of the conversation.

“You want to have a sci-fi vision and a short-term plan. The whole middle ground often fucks you up.”

“If you only have a short-term plan, you end up on this random walk where you can’t get out of your local maximum. And you can’t recruit someone by saying, ‘We’ve got a really good two-week plan and then we’ll figure it out.’”

But the opposite failure mode, a rigid multi-year roadmap, is just as dangerous. What works is holding a massive, almost absurd vision alongside a very concrete next step.

“We had this incredibly big vision: a new economic engine for culture, a new social contract, a renaissance. But that’s not a business idea. That’s a college-freshman-pot-smoking idea. So we also had this really concrete first step: make it dead simple to start a paid email newsletter. I don’t know exactly what all the steps are between ‘paid email newsletter’ and ‘insane cultural renaissance across the galaxy.’ But I know this is closer, and I can see the next couple of steps. When you have both of those ingredients, that’s how you can make progress.”

3) People hate stuff that sucks

On the topic of AI and creativity, Best argued that creators aren’t wrong to be annoyed:

“People actually hate stuff that sucks. And a lot of stuff that gets made with AI right now sucks.”

He illustrated the point with a story about a viral tweet.

“There was this video, really high production value, of an egg roll pooping out a spring roll into a deep fryer. Incredible technical thing. And the person tweeting about it is like, ‘Creative people, your days are over. The machine can do everything you do now.’ And if you’re somebody that makes movies, you look at that and think, ‘This person’s a fucking moron.’ And in their mind, the next step is: all this AI stuff is bullshit.”

“That’s obviously not true. We’re at the start of the most incredible technical revolution of my lifetime. It’s going to change everything. And it’s not going to substitute for taste. An egg roll pooping a taquito is not it.”

Best drew a parallel to his own experience watching AI coding tools improve, from barely functional a couple of years ago to genuinely transformative by late 2024. Something that’s 80% good enough is actually 0% good enough, he said. But when it hits 99%, everything changes.

4) Don’t hand control of your company to your critics

Substack’s stance on freedom of the press and freedom of speech has, at various points, made the company a lightning rod.

“We’ve taken a strong stand in favor of freedom of the press, freedom of speech. We think it is a crucial piece of that economic engine we’re trying to build. We’ve been through a number of moments where that’s been hard, where there’s somebody on Substack that people are like, ‘Why do you allow this? Why is this allowed to be here? Why don’t you just do what everybody else does and kick this off and get rid of it?’

“...We’ve been through left-wing censorship and right-wing censorship and we’ve kind of got now, eight years in or whatever, we’ve got this long track record that we can point to of being principled on this issue that I think not many others have.”

Best says they’ve weathered cycles where their stance was popular and cycles where it wasn’t, and that the political valence of the criticism has shifted more than once. His view: you need to know what you stand for before the pressure hits.

“There are moments where if we had bent from our principles, it actually would have killed us. If we started to hand over the moral authority, to say, ‘What do we think? How should Substack work?’ to whoever’s the loudest or most painful objector… giving that up would have handed control of the company to our critics. And it would have killed it.”

“The only way to make a company where nobody ever gets mad at you is to not succeed. It’s impossible to be number one at something, to be impactful, and not have some people hate you… You can’t define success as ‘no one is ever mad at us.’“

5) “I don’t want it in the abstract, but I do want it in the specific”

Best closed with a story about how he used to pitch Substack in its earliest days, and the reaction he kept getting.

“In the early days, I had this parlor trick. I would explain what the platform was: It’s a way that writers can go independent and make money on the internet. And people would be like, ‘Yeah, that sounds kind of cool, but nobody’s going to pay for that. I don’t want to pay money to get more email.’”

“And my parlor trick was: ‘Well, who’s your favorite writer?’ And they’d say, ‘It’s so-and-so.’ And I’d say, ‘Would you pay five bucks a month for them?’ And they’d say, ‘Yeah, for them I would, but that’s different. That’s somebody I really care about.’”

“You had this feeling of: ‘I don’t want it in the abstract, but I do want it in the specific.’ That’s a sign that you’re onto something.”

That instinct was validated early by Substack’s very first customer, Bill Bishop, who generated $100,000 in revenue within hours of launch. It gave Best and his co-founders the courage to keep building, even through the long stretches where growth was slower than expected.

“Professional gamblers often start their career with a winning streak,” Best joked. “We lucked out. We had this first customer that not only gave us the momentum, but gave us the courage of our convictions.”

Like this post? Forward it to your team! And for more weekly dives into the world of early stage startups, subscribe below.

You are receiving this newsletter since you opted in earlier; if you would like to opt out of future newsletters, you can unsubscribe immediately.

This newsletter is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. This newsletter may link to other websites and certain information contained herein has been obtained from third-party sources. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation.

References to any companies, securities, or digital assets are for illustrative purposes only and do not constitute an investment recommendation or offer to provide investment advisory services. Any references to companies are for illustrative purposes only; please see a16z.com/investments. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund which should be read in their entirety.) Past performance is not indicative of future results.

Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Content in this newsletter speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see a16z.com/disclosures for additional important details.

The Tech Week team just announced that submissions are now open for hosts looking to run an event at Tech Week Boston (5/26 to 5/31) and Tech Week New York (6/1 to 6/7). Apply here.

As the team behind Tech Week, we’ve grown the program to over 3,000 events across three cities. We’ve watched founders close enterprise deals, sign term sheets, and make their first hires all from a single well-run gathering.

After four years of running Tech Week—which started with just a few hundred events in LA in 2022, when marketing budgets across tech were tightening—we now have a pretty clear picture of what separates events that move the needle from ones that are just a line item on the marketing budget. Even as the economy has ebbed and flowed since then, budgets for Tech Week events have only increased, which tells you something about the ROI founders are seeing from this channel.

Most importantly for this audience, we believe that events should be a central part of your launch strategy. They are wildly underutilized by early-stage founders and represent some of the highest leverage go-to-market activity you could be doing.

Why events still matter

There’s a contrarian case to be made for in-person events right now. Every other marketing channel is getting noisier with the rise of LLMs. Social media is a knife fight in a back alley on a moonless night. You can use AI to send hundreds of emails, but recipients will just use AI to reply, “Please stop contacting me.” However, in person and live is a different type of magic. There is no way to automate a handshake, a demo over drinks, or the feeling a prospect gets when they meet the founder and decide they trust this person enough to buy from them.

Events are also the rare channel where the audience finds you. These types of conferences can allow you to decrease your acquisition costs because a potential lead has already flown cross-country for the week and is looking for things to do/learn. Most early-stage marketing is about defining a target persona, building a universe, and then paying through the nose to reach them. With events, especially during an industry gathering like Tech Week, you capture people you never would have targeted. Importantly, they are a different type of lead versus what you may purchase through an ad. They arrive at your event in a more open headspace versus someone who has been relentlessly hounded by your sales staff.

Most importantly for early-stage founders, events compress timelines. An enterprise deal that’s been stalling over email can close after a 45-minute conversation at a happy hour. A recruiting pipeline that took weeks of sourcing can materialize from a single fireside chat. And because event platforms like the one we offer for Tech Week collect LinkedIn profiles and contact details at RSVP, you walk away not just with the people who attended, but with a lead list of high-intent people who applied to attend as well. Time is your most precious resource in your first 12 months, so using events to get more done faster, is the most valuable outcome.

And don’t stress about cash, events are remarkably cheap if you’re resourceful. Several of the most successful events we’ve seen cost the founders almost nothing because they brought on sponsors to cover the tab. And events don’t even need to be expensive! It is about cultivating a vibe, something that cannot be purchased. Plus, at Tech Week, you can list your event for free!

Make the event your product

The thing you don’t want is to just host a happy hour with your banner on it. As an early stage founder, you need to design the event itself to be a demonstration of what you’re building. When someone experiences your product in a live setting rather than reading about it on a landing page, you skip the entire awareness-to-consideration funnel. They get it because they just lived it. Here are three approaches we’ve seen work exceptionally well.

Your event is your funnel. Sweatpals (SR004), a social fitness app from a previous speedrun cohort, co-hosted “Coffee & Chill” during New York Tech Week — a Saturday morning wellness event with cold plunges, a DJ, and networking, all inside the Classic Car Club in Manhattan. It was packed. And even crazier, people had to pay to get in. The Sweatpals team was able to demonstrate their product around community fitness, and not burn through piles of cash. Attendees walked away understanding the product because they’d just lived it – and that’s the shortest path from demo to acquisition.

Bake conversion into the event itself. The smartest founders make the event part of their sales funnel, not just the start of it. Sitch (SR005), an invite-only matchmaking app backed by speedrun, hosted the closing party for New York Tech Week. Entry was free, but only if you had a completed Sitch profile. Completed profiles were checked at the door. Every person who walked in was already an activated user, and Sitch promised to match attendees with someone afterward, turning the party into the first step of the product experience.

Bring a speaker who makes the marketing do itself. Northwood Space, founded by Bridgit Mendler, hosted an event during LA Tech Week on their literal factory floor (usually off limits) and talked about the next war being fought in space. Panelists were from all the heavy hitters: Anduril, Astranis, etc. and they had Ed Ludlow from Bloomberg moderate. This type of panel not only draws in attendees, it can also act as a form of external, organic marketing, with the event receiving coverage in Bloomberg.

The things everyone gets wrong

For this channel, the before and after details are what determine the event’s efficacy. Too often, we see founders pour all their effort into the actual party itself while neglecting the small things that will allow this to truly be a gamechanger for their business.

Before the event, invest in two things: a differentiated concept and a curated audience. There are hundreds of events happening during Tech Week in any given city. A generic mixer won’t cut it. Either lock in a speaker people want to hear from, or come up with a format that’s genuinely different. Then, put in the work on your invite list. The events that overperform are the ones where someone spent a week DMing people on LinkedIn and personally curating who would be in the room.

During the event, capture everything. You’ve spent weeks organizing a speaker, a venue, catering, and vibes—and then no one records the hot takes, no one posts a recap, and the whole thing dies the moment the last person walks out. Your event hits a hundred people max. Your content about the event should reach tens of thousands. Always, always, always be gathering content from the event to then share with the world.

After the event, follow up relentlessly. This is the single biggest missed opportunity we see. Planning an event is exhausting (ask us how we know) and by the time the event wraps up, founders tend to be drained and want to move on to the next thing. Don’t. The ROI of your event doubles with proper post-event execution: sales follow-ups, social media recaps, nurture emails to every RSVP who didn’t attend. This is the single most cultivated audience you will ever have! They just spent hours hanging out with your brand and product, so use that to your advantage. We highly recommend having the sales funnel ready to go before the event even starts. Come with a plan, leave with booked revenue.

How to use tech week

Tech Week is coming the last week of May starting with Boston and New York soon after (June 1-7), and event submissions are officially open!open. Listing your event is free and puts you on the same calendar as some of the biggest names in tech. If you’re building something and you want to get in front of the right people, submit your event here.

Have other questions about applying to speedrun? Reply and let us know. And for more weekly dives into the world of early stage startups, subscribe below.

You are receiving this newsletter since you opted in earlier; if you would like to opt out of future newsletters, you can unsubscribe immediately.

This newsletter is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. This newsletter may link to other websites and certain information contained herein has been obtained from third-party sources. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation.

References to any companies, securities, or digital assets are for illustrative purposes only and do not constitute an investment recommendation or offer to provide investment advisory services. Any references to companies are for illustrative purposes only; please see a16z.com/investments. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund which should be read in their entirety.) Past performance is not indicative of future results.

Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Content in this newsletter speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see a16z.com/disclosures for additional important details.

Over 19,000 startups applied to join the latest class of a16z speedrun, and just over 70 startups were ultimately selected.

Last week, those talented founders convened in San Francisco for the official kickoff of SR006. Today we’re bringing you a behind-the-scenes look at what the program is actually like for those participating.

Day 1 kicked off with a welcome session led by a16z speedrun GM Joshua Lu.

One big challenge for founders in a program like this is just knowing how to take advantage of all the services and operating team members at their disposal. The orientation introduced founders to the whole squad, covering everything from talent, GTM, brand, people, capital networks, and more.

After that, it was off to the atrium for a quick improv session and lunch.

Day 1 was all about getting to know the other founders in the program, so we gave everyone time to introduce themselves on stage. Bragging about themselves and their cofounders was encouraged. (It’s a skill you’ll need later when dealing with investors!)

Day 2 was when the rubber hits the road. The SR006 founders squeezed into big group sessions with their cohort leads to talk more formally about needs and where they’re at with their startups. Then it was back to the main hall for a fireside chat and Q&A with Substack CEO Chris Best, hosted by a16z GP Andrew Chen.

Founders like Best have seen both the ups and downs of startup life, and the Q&A session with speedrun founders afterward focused on lessons and watchouts applicable at the early stage in particular.

In between sessions, one big draw was the photography station with the speedrun logo set up outside. Founder photos were a must.

Day 3 brought the SR founders back together for a cohort office hours session about each team’s respective Demo Day goals, which was then capped off with an excellent fireside chat with Alt Capital’s Jack Altman, hosted by speedrun investor Marcus Segal.

After that, it was swag pickup time.

Then off to Google HQ for a few hours of games, drinks, and snacks.

And that was a wrap for week one. Over the next 11 weeks we’ll bring you more deep dives and inside looks at a16z speedrun as the amazing group of founders in SR006 builds toward Demo Day in San Francisco.

Have other questions about applying to speedrun? Reply and let us know. And for more weekly dives into the world of early stage startups, subscribe below.

You are receiving this newsletter since you opted in earlier; if you would like to opt out of future newsletters, you can unsubscribe immediately.

This newsletter is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. This newsletter may link to other websites and certain information contained herein has been obtained from third-party sources. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation.

References to any companies, securities, or digital assets are for illustrative purposes only and do not constitute an investment recommendation or offer to provide investment advisory services. Any references to companies are for illustrative purposes only; please see a16z.com/investments. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund which should be read in their entirety.) Past performance is not indicative of future results.

Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Content in this newsletter speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see a16z.com/disclosures for additional important details.

💡 Troy Kirwin says there are only three things that matter for a pre-PMF B2B founder to get out of their investor (beyond capital)

🤖 Fareed Mosavat says AI has made him fall in love with coding again

🔮 Jon Lai argues that most investors misunderstand the Genie 3 launch

💼 Join our talent network for more opportunities. If we see a fit for you, we'll intro you to relevant founders in the portfolio

Today’s essay is a guest post by Evan Armstrong, founder and CEO of The Leverage, a newsletter about shifts in the tech market.

Evan argues that the capabilities unlocked by tools like Claude Code, Cursor, and Codex call for a radically different approach to prioritization. We hope you enjoy it.

“Cocaine in the form of code” is how one founder recently described the experience of using Cursor to me. Another called Claude Code “his favorite video game.”

You’ve probably felt it right? There is something wildly addicting about coding agents—the sense of power is heady, each new command sends a rush of electricity to the fingertips, a feeling that you can make anything.

This is a trap.

The capability explosion in AI coding agents is real. Claude Code, Cursor, Codex—they’ve collapsed the time from idea to working software from weeks to hours. The skeptics were wrong. You can build internal tools, dashboards, workflow apps in an afternoon. It’s a new age for startups.

But the interesting question isn’t “can you?” It’s “should you?”

Build, Buy, or Delete

The classic question in operations is build versus buy. Do you take on the work yourself or outsource it to the market? Resistance to outsourcing has killed countless companies who insisted on vertical integration. The smart move is usually to own whatever “makes your beer taste better” aka differentiates your product, and buy everything else. You’ll never become a billion dollar company on the basis of your HR software, so just buy it from someone.

For early-stage startups, there’s a third option that’s almost always wiser: do without. Coding agents have made building so frictionless that we jump straight past the important question—does this process need to exist at all? Build-versus-buy assumes the thing should exist. Often it shouldn’t.

In Walter Isaacson’s biography, Elon Musk describes the algorithm he drills into SpaceX and Tesla teams:

1. Question every requirement. Every choice should have a real name attached to it. Then, you should question every single one of these assumptions, no matter who made it.

2. Delete any part of the process. Complexity is the death of speed and is anti-innovation. You should remove things to the point of failure. As Isaacson wrote, “If you do not end up adding back at least 10% of them, then you didn’t delete enough.”

3. Simplify and optimize. Only do this after deleting. Deletion is a precursor to improvement.

4. Accelerate cycle time. Again, speed is what matters. Increased speed greatly improves the flow of capital. But it should only follow step three, as there is no use accelerating processes that will end up being deleted.

5. Automate. Only once everything has been questioned, deleted, simplified, and accelerated can you automate. Musk learned a painful lesson when he tried to automate Tesla production too soon, leading to the “most hellacious period of his life” where he had to rip out expensive robots in favor of human labor.

The trap is that coding agents make Step 5 feel like Step 1. Building is so easy now that founders skip the prior questions. “Surely I can ditch my CRM? It feels like it is overpriced and isn’t customized to my needs. How hard can it be to make? After all, it’s just a database.”

(Famous last words.)

Instead, Elon’s algorithm would be asking: Can I delete this workflow entirely? Do I need a CRM right now or can I just make do with a Notion database? Can I tolerate a manual process? Is there something off-the-shelf that gets me 80% there for 30 bucks?

This is the danger of agentic coding. Making software cheap/easy to make does not mean you should use that capability replacing SaaS subscriptions willy-nilly (at least in the early stages). To understand why and what you should be doing, we return to Elon.

The Idiot Index for SaaS Has Inverted

Musk uses something he calls the “idiot index”—the ratio of a component’s cost to its raw materials. A $1,000 part made from $100 of aluminum has an idiot index of 10. “If the ratio is high,” Musk says, “you’re an idiot.” The delta is a waste of overcomplicated design, bloated process, or someone extracting margin from your ignorance.

For decades, the idiot index on software was brutal. Engineers were expensive. Code was hard to write, harder to maintain, and nearly impossible to deploy. The underlying primatives—logic, conditionals—were free. The finished product cost a fortune.

AI has started to invert this. So when code becomes cheap, what becomes expensive?

For early-stage founders, the answer is obvious. It’s your time.

Attention is the new raw material. Every hour in Claude Code building a custom CRM view is an hour not talking to customers, not closing deals, not iterating on the product that actually matters. The idiot index now measures how much founder time you’re burning on internal tooling that some Google Sheets and duct tape could handle for $7 a month.

And it’s not just the build. It’s everything after. The tool becomes a maintenance orphan, built by a coding agent and owned by no one. Then there are also the context-switching costs. Whoever you hire will be familiar with the basic productivity suite, or with tools like Notion or Airtable, but they will be baffled by whatever vibe coded alternative you came up with.

All of these add up to death by a thousand cuts, micro-distractions from doing the only thing that matters before your Series A: achieving product market fit. Everything else is pointless.

So before opening Claude Code, ask:

1. Does this need to exist? Many workflows don’t for the first year.

2. Can I do it manually at 70%? Good enough is good enough pre-PMF.

3. Is there a $30/month SaaS? The subscription is always cheaper than your time.

4. Who maintains this in six months? If the answer is “the founding team”, that’s a cost.

5. What else could I do with these two hours? If the answer is “talk to customers,” you’re done.

The founders talking to me about their addictions to coding agents feels like the perfect metaphor. Drugs give you nothing but dopamine, while coding agents give you dopamine and a deployment.

The feeling of progress plus the artifact of progress, minus the actual progress. That’s the worst trade a founder could make.

Like this post? Forward it to your team! And be sure to subscribe to Evan Armstrong’s newsletter, The Leverage.

For more weekly dives into the world of early stage startups, subscribe to our newsletter below.

You are receiving this newsletter since you opted in earlier; if you would like to opt out of future newsletters, you can unsubscribe immediately.

This newsletter is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. This newsletter may link to other websites and certain information contained herein has been obtained from third-party sources. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation.

References to any companies, securities, or digital assets are for illustrative purposes only and do not constitute an investment recommendation or offer to provide investment advisory services. Any references to companies are for illustrative purposes only; please see a16z.com/investments. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund which should be read in their entirety.) Past performance is not indicative of future results.

Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Content in this newsletter speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see a16z.com/disclosures for additional important details.

Our first double-feature is a guest essay from a16z speedrun investing partner Emily Bennet with pro tips for early stage founders thinking about how to model financials.

“We’re Too Early for a Financial Model.”

I hear this constantly, often from pre-seed or seed founders with the most impressive resumes. They have run scaled teams and built products used by millions, yet assume the early stage is too uncertain to justify time in Excel. Maybe you feel the same.

The logic sounds reasonable. Customers are still forming, pricing is provisional, and growth channels are unproven. Any model you build now will be wrong in months, perhaps weeks. You don’t want to create false precision or lose credibility with investors. So financial modeling gets pushed to later, once the business feels more “real.”

That intuition is understandable. I believe it is also wrong.

In my career as a venture investor, one of the strongest predictors of failure I have seen is a refusal to think structurally about an uncertain future. And one of the best ways to structure your view of the future is to build a financial model.

Here is why.

Your Fastest Way to Learn Is Through Testing

One of the most consistent patterns I’ve observed in startup failure is a failure to learn fast enough. Companies rarely fail because their initial assumptions are wrong. They die because they stay wrong for too long.

Research on forecasting shows that people are poor at predicting outcomes under uncertainty, but better at improving decisions when assumptions are made explicit and updated as new information arrives. These people don’t have some magical predictive powers. Their gift comes from structured iterations of their world view.

That is exactly the role a financial model should play at the pre-seed and seed stages. The goal is not to produce an accurate forecast, but to externalize beliefs in a way that makes them testable.

Building a model forces clarity about what must be true for the business to work. When reality diverges, the gap becomes legible. A CAC that lands at $120 instead of $40 is no longer a vague sense that growth is hard. It is a concrete signal about where learning is required.

Surprise is the mind killer

One of the biggest dangers of skipping a financial model is being surprised.

I have worked with many early-stage teams who were shocked when runway disappeared faster than expected, when they lacked the metrics needed to raise the next round at a reasonable valuation, or when their business could not sustain itself economically. A model does not eliminate these risks, but it makes them visible early enough to act.

Financial models accelerate learning in three ways.

1. They surface the real constraint, quickly revealing whether retention, pricing, sales efficiency, or burn rate is the bottleneck.

2. They create tight feedback loops. When reality diverges from the model, you immediately see where and by how much, turning gaps into concrete questions.

3. They enable intentional pivots by clarifying which assumptions broke and why.

In practice, the best founders treat their model as a living document. They update it regularly, use it to set near-term goals, and communicate progress clearly. You do not need a three-statement, GAAP-perfect model. A single spreadsheet with key assumptions, burn, runway, and a few scenarios is enough, as long as the logic is clear and evolves.

The Three Things You Actually Need to Track

Seed-stage startups should focus on three key tiers of signals that answer the question: can this company survive, make progress, and ultimately work as a business?

Tier 1: Financial Viability

This is the baseline. You need a clear view of cash in, cash out, burn, and runway over the next 12 months. The goal is to eliminate surprises about survival.

Tier 2: Structural Goal-Setting

This is your OKR layer. Which outcomes signal real progress toward product-market fit? What must be true of the business for you to believe you are building something users genuinely want?

For a consumer company that is not yet monetizing, or a very early B2B company, the P&L may show only outflows. You still need a quantified north star. For example: “We will grow weekly active users by X percent month over month while maintaining Y percent retention in our core cohort.”

That metric matters less as an external narrative and more as an internal truth. It becomes the scorecard for whether the product is resonating with users. Consistent goal tracking also creates a baseline for comparison, allowing you to benchmark against similar companies and distinguish surface-level activity from real, competitive traction.

Tier 3: Business Model Efficiency

Finally, you need to pressure-test whether the business model can work at all. What does steady-state unit economics look like? Is there a path to durable, scalable ROI?

You will likely not know true LTV at seed, and that is okay. What you need is a defensible hypothesis. “We believe LTV will be roughly $500 based on comparable products, which implies CAC must stay below $150.” From there, you track early indicators such as cohort retention, payback, and expansion to see whether reality is converging toward that hypothesis or drifting away from it.

Where to start

Still, I know you have a million things to do. So, I wanted to make this as easy as possible. To help I created three interactive trackers to kickstart deeper visibility into the early metrics that matter:

Financial Viability Tracker - A straightforward cash tracking tool that answers the fundamental question: “Will we still have money at the end of the year?” It includes:

• Revenue, fixed, and variable cost inputs

• Current cash, monthly burn, and runway calculation

Structural Goal Setting Tracker - An OKR-focused tool that helps founders work backward from their next fundraise. It includes:

• North star metric tracking with growth projections

• Multiple objectives (OKRs) with progress tracking

Unit Economics Tracker: A tool for pressure-testing business model viability. It includes:

• LTV:CAC ratio calculations

• Cohort retention tracking

So yes, don’t reject the push to build a model. And use this as a starting point! You’ll have a much better grasp of your business as it naturally evolves.

Have other questions for Emily? Reply and let us know. And for more weekly dives into the world of early stage startups, subscribe below.

You are receiving this newsletter since you opted in earlier; if you would like to opt out of future newsletters, you can unsubscribe immediately.

This newsletter is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. You should consult your own advisers as to those matters. This newsletter may link to other websites and certain information contained herein has been obtained from third-party sources. While taken from sources believed to be reliable, a16z has not independently verified such information and makes no representations about the enduring accuracy of the information or its appropriateness for a given situation.

References to any companies, securities, or digital assets are for illustrative purposes only and do not constitute an investment recommendation or offer to provide investment advisory services. Any references to companies are for illustrative purposes only; please see a16z.com/investments. Furthermore, this content is not directed at nor intended for use by any investors or prospective investors, and may not under any circumstances be relied upon when making a decision to invest in any fund managed by a16z. (An offering to invest in an a16z fund will be made only by the private placement memorandum, subscription agreement, and other relevant documentation of any such fund which should be read in their entirety.) Past performance is not indicative of future results.

Charts and graphs provided within are for informational purposes solely and should not be relied upon when making any investment decision. Content in this newsletter speaks only as of the date indicated. Any projections, estimates, forecasts, targets, prospects and/or opinions expressed in these materials are subject to change without notice and may differ or be contrary to opinions expressed by others. Please see a16z.com/disclosures for additional important details.