Confidence as a Founder: Where Conviction Comes From

TL;DR: Confidence as a founder is not a personality trait — it's the accumulated result of small bets that paid off. Based on 200+ founder interviews on the PMF Show, founders consistently describe conviction arriving after the first decisions worked, not before. Start with a problem you'd pursue regardless of outcome, then let evidence build confidence.

After interviewing 200+ founders on the PMF Show, the most consistent finding about confidence as a founder is how few of them had it at the start. Repeatedly, the founders with the most impressive track records describe the early years as scary and uncertain — and describe their confidence as something they earned retroactively. This article covers where that confidence actually comes from, how founders held conviction while their products were failing, and where confidence tips over into a liability.

What actually creates confidence as a founder?

Evidence, compounding. Paul Mikesell, founder of Carbon Robotics, has been at five startups and through four IPOs — a track record that would suggest natural self-assurance. He describes it as a learned loop instead.

"I don't think I'm special in my ability to choose good companies. I'm, I may be special in my ability to have conviction." — Paul Mikesell, Carbon Robotics

According to Paul Mikesell, founder and CEO of Carbon Robotics, the mechanism is straightforward:

"You have conviction around your beliefs, right? It's like I can identify what I believe in and then activate on that and put my whole time into it. Being able to do that can be challenging and scary. But once you kind of, once some of those earlier bets start to pay off, you realize you've made good decisions and you start to trust yourself a little more." — Paul Mikesell, Carbon Robotics

Note the order: act first, then trust yourself. Across the 200+ interviews, this sequence is nearly universal. Founders describe the first period as "challenging and scary" and locate their confidence in the outcomes of decisions two and three, not decision one.

Key stat: Five startups and four IPOs, and Mikesell still attributes the result to conviction rather than picking ability.

Is confidence the same as conviction?

No — and the distinction matters, because they come from different places. Confidence is a belief about yourself. Conviction is a belief about the problem, and it's far more durable under failure. Arvind Jain, co-founder and CEO of Glean, made the difference explicit when asked how he stayed committed to a two-year build with no market feedback.

"Part of it is actually what is the goal that this founder has. Is the goal to actually build a successful company or is the goal to solve a problem, something that you're passionate about? For me, it was the latter. I was not trying to get a win. I was actually trying to solve that problem." — Arvind Jain, Glean

That framing is load-bearing. If your motivation is winning, every month without traction is direct evidence against you, and confidence erodes on a schedule. If your motivation is the problem, months without traction are just information about the problem.

As discussed on the PMF Show, this is also what lets founders go against prevailing methodology. Glean's approach — build for two years with a strong internal vision — runs directly counter to the Lean Startup orthodoxy most founders run by default. Conviction about the problem is what made that defensible.

Key stat: Glean's founder committed to a roughly two-year build before broad market validation.

How do you stay confident when the product isn't working?

You often don't — and several founders on the show say that's the correct response. Jordan Dearsley, co-founder of Vapi, went through a sequence of pivots before landing on voice AI, and his YC group partner predicted the pattern exactly.

"Michael Seibel. He was one of our group partners at YC. He told us quite early on, I'm skeptical that you guys are going to continue doing what you're doing for more than three months if it isn't working after three months. And he was right every single time. And every three months we would pivot, we would pivot, we would pivot." — Jordan Dearsley, Vapi

What Dearsley identifies as the missing ingredient isn't confidence — it's attachment to the problem.

"I think what also needs to be there is, like, some level of passion or determination for the problem... I think it does take some sort of like personal connection to a problem or at least some like long-term conviction just to hang in there long enough to see it through to something that's meaningful." — Jordan Dearsley, Vapi

He's blunt about the earlier version. With their calendar app, the team simply executed whatever users asked for next — calendar links, then scheduling, then email — with no long-term view. His verdict: "So, of course, we were never going to achieve something big long-term."

Vapi's financial discipline supported the search. The team was default alive and roughly breakeven, having spent about $1M over three years, and even returned $500K of investor money to buy freedom to explore.

Key stat: Vapi pivoted roughly every three months, spent ~$1M over three years, and returned $500K to stay default alive.

Where does confidence come from when the market disagrees with you?

From pattern recognition about technology curves rather than from current evidence. Surojit Chatterjee, founder and CEO of Ema, started building on AI when the technology visibly didn't work.

"At that time, a lot of things did not work. So you have to have a belief in where this technology will be in a few years and start building with that future." — Surojit Chatterjee, Ema

He'd been in this exact position before. In 2007 he launched mobile advertising at Google and couldn't convince his own colleagues:

"When I was starting mobile advertising at Google in 2007, I struggled to convince even internally this will be big, forget externally... He's like, I like your enthusiasm, but I don't think it'll be true." — Surojit Chatterjee, Ema

Chatterjee's framework distinguishes between two things that look identical from outside: a product that doesn't work, and a product that doesn't work yet on a trend line that's clearly moving. His words were that the technology was "not working, very broken, but the trends were there." Confidence, in his telling, is calibrated to trend direction rather than to current state.

Key stat: Chatterjee had already been early once — mobile ads at Google in 2007, roughly a decade before mobile dominated ad spend.

Can you be too confident as a founder?

Yes, and the risk rises with success. Saurav Chopra, co-founder of Perkbox, described how a first exit changed his relationship with confidence on his second company.

"As a first time founder, or if you had an exit that was perhaps not so great, the financial motivation is obviously very, very high. But when you do a second round around, it's not really about the financial motivation... So that's the positive that you can really think about the long-term and you're not driven by, hey, quarter to quarter." — Saurav Chopra, Perkbox

But the same experience produced a specific failure mode:

"I can reflect and say, huh, maybe I was a bit more overconfident in this one. Maybe I should have tested this a bit more." — Saurav Chopra, Perkbox

The practical test is whether your confidence is reducing the amount of evidence you gather. Conviction that makes you persist through a hard build is productive. Conviction that makes you skip customer validation because you're sure you already know the answer is the version that costs a year.

Key stat: Chopra's second-company reflection: overconfidence caused him to under-test assumptions he'd have validated as a first-timer.

What if you don't feel confident enough to start?

Then find someone who is — the confidence doesn't have to be yours. Rick Song, co-founder and CEO of Persona, is candid that the company wouldn't exist without his co-founder.

"He was the one who, from a timing perspective... I'm not the type to jump. He's the one I oftentimes credit to actually being the entire impetus of why the startup exists at all. Because I would have never jumped." — Rick Song, Persona

Song also reframes the downside honestly. For an experienced engineer, the realistic worst case of a failed startup is "a year or two sabbatical and you come back" — a materially smaller risk than the one anxiety tends to model.

His broader observation about differentiation is worth sitting with:

"By default if you want to be that one percent, you have to think about something differently. But I think it's funny when... 99 percent of founders end up thinking exactly what the other 99 percent are thinking. So there's by default, no differentiation in the thought process." — Rick Song, Persona

The uncomfortable implication: if your idea feels obviously sensible to everyone you describe it to, that comfort is not the same as confidence being warranted.

Key stat: Persona's CEO says the company would not exist without a co-founder willing to move first.

Key Takeaways: Building Real Confidence as a Founder

1. Confidence follows action, not the reverse. Paul Mikesell describes trusting himself only after early bets paid off — five startups and four IPOs later, he still frames it that way. 2. Conviction about the problem outlasts confidence in yourself. Glean's Arvind Jain committed to solving a problem rather than getting a win, which is what made a two-year build survivable. 3. Pivoting is not a confidence failure. Vapi pivoted roughly every three months, exactly as their YC partner predicted, and that iteration is what found voice AI. 4. Default alive buys you the freedom to keep looking. Vapi spent ~$1M over three years and returned $500K rather than accept pressure that would narrow their search. 5. Calibrate to the trend line, not the current state. Surojit Chatterjee built Ema when the technology was "very broken, but the trends were there" — the same read he made on mobile ads at Google in 2007. 6. Watch for confidence that reduces your evidence. Perkbox's Saurav Chopra identifies overconfidence on his second company as the reason he under-tested assumptions. 7. Borrowed conviction is legitimate. Rick Song credits his co-founder as the entire reason Persona exists; he says he would never have jumped alone. 8. Consensus comfort is not confidence. If 99% of founders would think what you're thinking, differentiation isn't coming from your thesis.

FAQ: Common Questions About Confidence as a Founder

Q: How do founders build confidence as a founder in the early days?

A: Almost entirely through accumulated small wins. Founders on the PMF Show consistently describe the first stretch as scary and describe their confidence as arriving after the second or third decision worked, not before the first one.

Q: What's the difference between founder confidence and conviction?

A: Confidence is belief in yourself; conviction is belief in the problem. Conviction is more durable because failure becomes information about the problem rather than evidence against you.

Q: Is it normal to lose confidence when a startup isn't working?

A: It's the common experience, not the exception. Vapi pivoted every three months for a stretch; their YC group partner had predicted exactly that. What sustained them was attachment to a problem, not sustained self-belief.

Q: Can a founder be overconfident?

A: Yes, and the risk increases after a successful exit. Perkbox's Saurav Chopra says his main second-company mistake was under-testing assumptions because he was more confident than the evidence justified.

Q: What if I'm not confident enough to start a company?

A: Many founders weren't. Persona's Rick Song says he would never have left his job without a co-founder who was ready to move — and notes the realistic downside for most experienced operators is closer to a one- or two-year detour than a career-ending one.

Sources: Listen to the Full Founder Stories

• Paul Mikesell, Carbon Robotics — on conviction as a learnable skill, across five startups and four IPOs.
• Arvind Jain, Glean — on choosing the problem over the win, and committing to a two-year build against Lean Startup orthodoxy.
• Jordan Dearsley, Vapi — on pivoting every three months, returning $500K, and why long-term conviction beats short-term responsiveness.
• Surojit Chatterjee, Ema — on building for where technology will be, and being early to mobile advertising at Google in 2007.
• Saurav Chopra, Perkbox — on how a successful exit changed his motivation and produced overconfidence the second time around.
• Rick Song, Persona — on borrowed conviction from a co-founder and why consensus thinking produces no differentiation.

Last updated: August 2026