How to Pivot Well
Investors from the speedrun team share the advice they give founders considering a pivot
This Week’s Big Ideas 💡
❓ Emily Bennett shares one of her favorite cofounder diligence questions
💰 Fearn (SR005) announced they’ve raised a $5.5M seed to help inventors file high-quality patents in mere minutes.
☀️ Orbital (SR007) raised a $5M pre-seed to build solar-powered intelligence
💼 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 desperate founder, betting their last dollar on a pivot, leans on the lore: Slack was a game studio. Instagram was a check-in app. Twitter was a podcast directory. These iconic stories of the web 2.0 era give all founders hope that unimaginable riches are just one more iteration away.
But are these stories even relevant anymore? AI model capabilities shift every quarter, which means that a new pivot is theoretically possible every quarter. Should every company be pivoting accordingly? Knowing when to pivot, how fast, and toward what has quietly become a core founder skill in the AI era.
So we asked speedrun investors Emily Bennett and Troy Kirwin what advice they give to founders considering a pivot.
First, what a pivot actually is
In basketball, a pivot means one foot stays planted. Move both feet and it’s traveling. (Go Knicks.) Startups work the same way. A good pivot keeps something planted, oftentimes the team and the investor’s cash, while everything else moves. The term originated from author Eric Ries and his 2011 book A Lean Startup, though he initially blogged about it in 2009.
Ask a speedrun investor and they’ll tell you that sometimes the worst failure comes due to refusing to pivot at all. As Emily Bennett puts it:
“Pivots are really hard to do. But what I always say is that the worst thing to do is to run with obstinate determination at the wrong thing, once it’s already become very clear that it’s the wrong thing.”
There’s even a case that the founders best built to pivot are the ones with the least to unlearn. Troy Kirwin argues that inexperience can be a feature:
“One of the ways young, first-time technical founders are empowered versus more senior founders is that they don’t come with any experience. So they’re not backed into a corner about what industries they feel they need to build for. They can be more of a heat-seeking missile to go test and validate any market. That’s the double-edged sword of founder-market fit.”
Three recent pivots worth studying
Clay: pivot toward the heat. Kareem Amin co-founded Clay in 2017 with a beautiful, somewhat abstract mission to make the power of programming accessible to everyone. The team’s product was a spreadsheet that could talk to the whole internet. Five years later, revenue was still close to zero. (Ouch.) The handful of customers they did have were using it for everything from accounting to sales. The unlock was noticing that some of their most active users were growth teams chaining data sources together to find customers. So, Clay rebuilt its entire identity around go-to-market. Two years after that pivot they hit $30M ARR. By early 2026 they were at $100M+ ARR.
Lovable: pivot toward insight. In June 2023, Anton Osika spent a weekend building a command-line tool that generated codebases from plain English. He then open-sourced the project and it pulled 50,000 GitHub stars in about two months. The beacons were lit; customers were calling for product. He and co-founder Fabian Hedin rebuilt the idea as Lovable, a closed, polished tool for non-technical builders. It hit $100M ARR in roughly eight months. Plenty of teams have turned repos into companies, Databricks sold Spark back to the same engineers already running it, and LangChain, Ollama, and ComfyUI are monetizing their own user bases right now. Lovable ran a stranger play. While the repo proved there was demand, the founders realized that the real market would be the 99% who couldn’t use a command-line tool at all.
Cursor: pivot toward yourself. Before Cursor, Anysphere’s four founders spent nearly a year building AI autocomplete for CAD software. CEO Michael Truell calls that period “wandering in the desert,” and his diagnosis of what went wrong is sorta obvious. They weren’t mechanical engineers, they weren’t passionate about the space, and there wasn’t much CAD data to train on. That idea might be a good one, but they weren’t the founders to build it. So, they swung to a market where they were the customer, where they actually cared, and where the internet was full of training data—aka coding.
The playbook
Across these stories and our investors’ answers, the same principles keep showing up.
When should you pivot?
Emily warns is that early stage founders sometimes fall in love with their own idea that they fail to notice the lack of market pull for it:
“Very early in the founder journey, you’re instinctually looking for market pull. You’re looking for signal from users that there’s real, unrestrained demand for what you’re pitching them… a commercial appetite underlying it, a willingness to pay. I’ve seen some founders get so excited by their concept, by the market they think they’re building in, that they’re blind to the lack of signal on that front.”
And when the signal isn’t there, she pushes founders past the comfortable explanation:
“You can either say, well, they’re not getting it. Or, the more likely answer, is that there’s something you’re not getting.”
Troy turns this into a cheap, time-boxed test that forces the truth out fast:
“For founders who are thinking about pivoting, I say: Why don’t you go try to fill your calendar with another fifty conversations with customers over the next two weeks? If you can’t even get those conversations, then that itself is a signal… If it feels like pulling teeth to get them to hop on another call, or they ghost you, then that’s a signal. You’re never going to have one hundred percent conversion, but if you can’t get fifteen or twenty percent, then that’s a problem.”
Troy says that sometimes fear can keep founders from acting on the signals they’ve gotten from the market:
“Folks are sometimes afraid of the truth, because the truth can lead toward sunk-cost fallacy and realizing you just burned the last three to six months on something you now need to shutter… But the most successful founders are truth-seekers.”
How should you pivot?
Remember the basketball rule: a good pivot keeps one foot planted. What stays planted, Emily argues, is everything you already learned, including the pile of conversations with customers a founder is tempted to write off as wasted:
“Look at all the conversations you just had that you now may think are totally moot. You’re not wasting time. Nothing is wasted. Look at the prior concept you were trying to launch, and the main positive insights you can parse from that experience then become the foundations to enable you to move faster at a new idea. Usually there’s a kernel of something worth keeping.”
From there, Emily runs a three-step playbook. It starts deliberately small:
“Find ten people who are willing to go incredibly deep with you and share all of their thoughts. Make that a representative pool… really use their input to inform your new market direction.”
Only once that depth of insight is in hand does she widen the funnel, running side-by-side A/B tests on messaging and archetypes to layer quantitative signal on top of the qualitative.
As for the conversations themselves, Troy says the goal is to learn:
“This is Sales 101 stuff. You’re not necessarily pitching on the call, but doing discovery on the call. State your hypotheses of a problem you’ve identified with other customers and ask if that resonates.”
And in 2026, there’s no excuse to delay the test. Emily’s point lands especially hard given what Lovable became:
“Now it’s so easy to build, you can literally pull up Lovable and create a rapid, interactive prototype and put that into the world. So your excuse should never be that you need to do product debt in order to get customer feedback.”
Anything else founders should think about before pivoting?
The hardest pivots tend to be for companies that aren’t clearly dead, but aren’t growing fast enough. Here Emily makes founders confront whether they’re building a venture-shaped company:
“There are good businesses that don’t grow unbelievably rapidly. They’re good lifestyle businesses, they’re actually profitable. But it just means you probably aren’t going to be able to command scaled venture dollars… It all goes back to the venture business model.”
For an idea that truly isn’t working, though, she thinks founders have the fear backwards:
“What’s scarier is building a product for nobody and wasting a lot of money.”
And for anyone dreading the conversation with their investors, Troy offers the reassurance that a pivot is a green flag, not a red one:
“Sometimes founders are nervous to come to us with the potential of a pivot. And to us, sometimes I’m excited to hear that, because it just means they’re moving fast. Getting to conviction that a space isn’t working is as strong a sign as showing that it is working. To me, it’s a sign of a great team.”
That’s it for this week! Thanks to Troy and Emily from our investing team for sharing their insights.
For more weekly dives into the world of early stage startups, subscribe below.
The Cursor "pivot toward yourself" point is the keeper: building where you're already the customer fixes motivation, training data, and discovery in one move.
Troy,
I wanted to ask you something. How can someone realistically conduct those conversations when they're operating in an environment where there is zero connectivity at night?
I'd genuinely value your perspective on this.
Thanks.