There's a number that greets every would-be founder before they've written a line of code or made a single sale: nine out of ten startups fail. It shows up in accelerator slides, pitch-deck disclaimers, and the quiet voice at 2 a.m. asking who you think you are. And it does real damage — because it convinces thousands of capable people that the odds are hopeless before they even check whether the odds apply to them. The truth, per the 2026 data, is both more nuanced and far more encouraging.

That 90% figure comes from the Startup Genome study — and it measures a very specific thing: venture-backed startups failing to deliver 10x venture returns. By that definition, a profitable company doing $5 million a year is a "failure" if it raised a big Series A. But most people building something aren't playing the venture-returns game. They're trying to build something real that pays them and lasts. For them, the honest number is the BLS one: about 20% of all businesses close in year one, roughly 50% by year five. Hard, but nowhere near hopeless. This is exactly the distinction Michael Dermer has spent years making — that the story we tell founders about their odds shapes whether they ever start at all.

The "90% fail" number isn't a fact you need to accept — it's a game you may not even be playing. Most founders aren't trying to be a unicorn. They're trying to not go broke. Those are completely different odds.
— Michael Dermer

So the first act of building isn't hustle. It's knowing which game you're in — and then knowing what actually kills the businesses that don't make it. Because the cause of death, it turns out, is remarkably consistent.

Why "just build it" is the most expensive advice in startups

Here's the trap. The founder instinct is to treat building as the point — get the product out, ship fast, figure out demand later. It feels like momentum. The data says it's the single most expensive mistake a founder can make. When CB Insights studied 431 failed venture-backed companies, the headline finding wasn't that they lacked money — those 431 companies had raised a combined $17.5 billion, with a median of $11 million each. They had money. What they lacked was evidence that anyone wanted what they were building.

That's why the top-cited cause of failure — "ran out of capital," at 70% — is misleading. Running out of cash is almost never the root problem; it's the final symptom of the real one: poor product-market fit (43%). The Startup Genome data drives it home: 74% of failed startups scaled prematurely — hiring and spending on marketing before confirming that anyone wanted the product. The move that separates survivors from statistics isn't working harder or raising more. It's validating demand before building, the discipline that repeat founders learn the expensive way and first-timers can simply borrow.

These companies didn't die because they ran out of money. They ran out of money because they never confirmed anyone wanted what they were building. That's a fixable mistake — but only before you make it.
— Michael Dermer

So if cash is the symptom and fit is the disease, the next chart is the one every founder should sit with — because it separates what kills you from what you can actually control.

Why the odds feel lonelier than they are

Sit with the founder-experience data and something quietly hopeful emerges. First-time founders succeed about 18% of the time; repeat founders who've succeeded before hit 30%. That 12-point gap is real — but here's the part that matters: a prior failure barely moves the needle at all (20% versus 18%). Experience isn't magic, and failure isn't a scarlet letter. What repeat founders actually carry forward is a single learned discipline — they stop assuming demand and start testing it. And that discipline is completely transferable. You don't need to fail expensively to learn it; you can borrow it on day one.

This is where the numbers meet the reason The Lonely Entrepreneur exists. The 90%-fail myth doesn't just misinform — it isolates. It convinces every first-time founder that they're staring down odds that everyone else somehow beats, that their doubt is a signal they're not cut out for this. The data says the opposite: nearly everyone struggles with the same handful of problems, most of them solvable, and the founders who make it aren't the ones who felt no fear — they're the ones who didn't face it alone. When a founder learns that the "brilliant" repeat founder mostly just knew to validate first, the intimidation drops away. The gap between making it and becoming a statistic is smaller, and more learnable, than the mythology admits.

A prior failure barely changes your odds — which means failing once isn't the end of anything. What actually separates the founders who make it isn't talent or luck. It's whether they had someone to learn the hard lessons from before they cost everything.
— Michael Dermer

Which raises the practical question: if the odds are more beatable than they seem, what does actually beating them look like? The founder-type data points straight at the answer.

How to beat the odds — starting before you build

The most encouraging thing about the failure data is that the highest-leverage move costs almost nothing and happens before you commit real money. The number-one root cause of failure — poor product-market fit (43%) — is the one cause you can fully address in advance. So the first move is simple and hard: validate demand before you build. Talk to real potential customers, check search volume, look at whether competitors are actually making money. The founders who skip this aren't braver; they're just paying to learn a lesson they could have gotten for free.

The second move is to resist the premature-scale trap. Startup Genome found that 74% of failed startups scaled too early — hiring, spending, and building teams before confirming anyone wanted the product. Stay lean until the evidence says otherwise; a small business with proof of demand beats a well-funded one running on hope. And the third move is the one the data can't show you but every founder feels: don't do it in isolation. The single learnable edge repeat founders have is transferable, which means the fastest way to close the experience gap is to borrow it — from a mentor, a peer group, a community of people who've made the mistakes you're about to make. That's not a soft nicety. Given that a prior failure barely dents your odds, the real risk isn't failing; it's failing alone with no one to help you extract the lesson and go again.

It's not enough to tell a founder "the odds are better than you think" — that's just a nicer statistic. You hand them the one discipline that separates winners from statistics, and a room of people who've already made the mistakes, and suddenly the odds are genuinely theirs to beat.
— Michael Dermer

The failure mode here isn't a lack of talent — first-time founders are among the most driven people alive. It's isolation disguised as independence: believing you have to figure it all out yourself, when the data shows the whole difference is a lesson someone else already learned. Effort spent building in the dark burns the runway. Effort spent validating first and building with support is what puts you in the 18% — and then, over time, moves you toward the 30%.

What the survival data is really measuring

Zoom out from the percentages and the numbers are measuring something the startup world rarely admits: that the odds every founder quotes are averages across everyone — including the majority who did zero validation and built entirely on assumption. When CB Insights notes that the median failed company was "walking dead" for three-plus years before officially closing, and that most founders knew something was wrong long before they admitted it, you're not looking at a story about bad luck. You're looking at a story about isolation — founders too alone, or too proud, to hear the truth early enough to act on it.

And here's the reframe that matters: the survival statistics aren't a verdict on whether you're good enough. They're a map of the mistakes that are avoidable and the ones that aren't. Bad timing (29%) is largely outside your control; poor product-market fit (43%) is almost entirely inside it. A world where every first-time founder knew that the odds are more beatable than the mythology claims, validated before building, and did it with people around them instead of alone isn't just a kinder world for founders — it's one with more businesses that survive, because more of the people building them make the one move that actually matters. The odds of entrepreneurship will always be humbling. The loneliness of facing them doesn't have to be part of the deal. That's the whole reason this company has a name.

Building a business, and facing the possibility of failure, can take a real toll. If you're personally dealing with chronic stress, burnout, anxiety, or thoughts of self-harm, please reach out to a mental-health professional or a trusted person in your life — you don't have to carry it alone. In the US, you can call or text 988 to reach the Suicide & Crisis Lifeline, 24/7.