Why startups fail with a good product

Most post-mortems blame the product. The team shipped something nobody wanted, and the market returned its verdict. That happens. But there's a more common and more frustrating version: the product is genuinely good, the team is capable, the work is real — and it still stops.

I know this one from the inside.

How it started

A year and a half ago, my co-founder and I were scrolling Instagram. We kept passing pages running events around healthier living — a whole small economy of them. He said we should start a page like that. Something in the lifestyle space.

I said we should build the app instead.

He looked at me like I was joking. I wasn't.

We started with nothing. I didn't know what UI or UX meant. I didn't know a single programming language. Our learning materials were the Steve Jobs film, the Zuckerberg one, and ChatGPT. We produced a genuinely ridiculous AI-generated interface. But we worked on it every day — seven months, six hours a day — and I organised the two of us as if we were a thousand-person company.

Some of what we designed was right. Within about a year, versions of the same ideas appeared inside much larger platforms. So the instinct wasn't the problem.

The product wasn't the problem either. We finished it, and everything stopped anyway.

The actual failure: building the end state directly

Here's the mistake, stated plainly. We were trying to build the finished vision in one move.

A mature product is the result of a long sequence — early users, feedback, distribution, revenue, iteration. Each stage teaches you what the next one should be. When you skip to the end state, you build something with no way to enter the world. There's no smaller version for someone to try, no first customer who needs exactly that slice, no revenue to fund the next step.

It isn't a product failure. It's a sequencing failure.

The second mistake compounded it: our plan was to build the app, send it to ten investors, and expect one to fund it. Look at that honestly. What investor funds a first-time team with no traction and no revenue? And what inexperienced founder should accept that money if offered? Neither half made sense — but it let us avoid the harder question of how the thing would actually reach anyone.

How to tell if this is happening to you

A few signs worth checking:

• Your roadmap describes a finished product, but you can't name the one feature a real person would pay for next month.
• You can describe your users in general terms but haven't spoken to ten specific ones.
• Your funding plan requires someone to believe in the vision, because there's nothing yet to measure.
• Most of your energy goes to building and raising, and very little to distribution — to how anyone actually finds this.

If several are true, the risk isn't your product quality. It's that you have no path from here to a first user.

What the failed version was actually worth

Here's what I believe now, having done it: an ambitious idea may have a low probability of succeeding, but the knowledge it hands you is worth more than years spent inside a large organisation doing the safe version of the same work.

Think about cooking. Say you make a genuinely good burger. You can go work the line at McDonald's, or you can decide to compete with McDonald's. Obviously the second won't work — not at your size, not now. But if you actually try, you stop being someone who makes burgers. You become someone who understands the whole system: cost, supply, positioning, why people choose one thing over another. That knowledge follows you everywhere afterwards.

I don't like textbooks, and this is why. A thousand pages to teach what a few pages plus real experience teaches better and faster.

What changed

When we reopened the project, I called a meeting and said what I'd come to believe: the risk is too high, because we're building the end state directly. This has to start from something smaller and more fundamental and evolve toward what we're imagining.

So we inverted it. The larger vision stayed, but a smaller, sharper product went first — one that could reach real users, generate real feedback, and fund what came next.

The bigger shift was where my attention goes. Before, nearly all of it went to innovation and finding an investor. Now roughly seventy per cent goes to infrastructure and reducing risk: the landing page, content that works across channels, understanding the market and the customer, driving costs down.

None of it is glamorous. All of it has produced results.

A good product that can't reach anyone isn't a good product yet. It's a good prototype waiting for a structure around it. The ambitious project never launched — but it built the person who could launch the next one.

If you've built something that hasn't found its way to users, that's usually a structure problem rather than a product problem — and it's the kind of problem I work on.