There’s really no such thing as ‘vision without a business model.’ Instead, there are two visions: the one founders share with investors, and the one shown by the company’s numbers. What people call a vision failure is usually just the gap between these two, left unaddressed until it becomes obvious.
The Story the Market Tells Itself
Over the past decade, many people have come to believe that visionary founders are risky, while operator-founders are safer. The evidence seems convincing: ten major startups from 2019 to 2025, like WeWork, Theranos, Bolt, Fast, Olive AI, Hopin, Humane, and Convoy, all raised huge sums based on their stories. And all of them failed.
The market took this to mean that vision is risky and discipline is safe. Andreessen’s ‘software is eating the world’ encouraged big stories, while Sequoia’s ‘Adapting to Endure’ memo signaled a shift. Since 2022, every accelerator pitch deck now includes a unit economics slide that wasn’t there before.
This shift isn’t wrong, but it’s not the whole story.
.jpg){kind=link}
What Doesn’t Fit That Frame
When I look at what really happened with those ten failed companies, the pattern isn’t what people think. The vision didn’t fail. The problem was with disclosure.
In every case, the real numbers, like contribution margin, were never shared publicly. This wasn’t always because the numbers were bad, though sometimes they were. The real issue was a lack of transparency. For example, Humane raised $240 million but only made $9 million in sales before a Verge review exposed the numbers. Fast was burning $10 million a month while making just $600,000 a year, a huge gap. Olive AI raised $902 million and was valued at $4 billion, but hospitals found its setup process so complicated that making a profit was nearly impossible.
Every turning point came from outside pressure: a Wall Street Journal exposé, an IPO filing, an independent rating, a Forbes investigation, or a lawsuit. Visionary stories last until a formal document reveals the real numbers and brings the truth out.
What people often miss in the ‘vision is bad’ story is that these founders weren’t always wrong about the future. The real problem was that their internal numbers didn’t match what investors saw. The vision itself wasn’t the issue, it was the lack of transparency that eventually failed.
Two Visions, One Clock
After years of building, advising, and watching companies in Dubai, Singapore, London, and Frankfurt, I’ve come to believe the real issue isn’t vision versus business model. It’s about whether a company is disciplined or fragile in its disclosures.
Every company has two visions at the same time. One is in the pitch deck, board meetings, press, and all-hands calls. The other is in the numbers, like contribution margin, burn rate, retention, and CAC payback. In strong companies, these two match up. In struggling ones, they drift apart until someone outside forces them to align.
The gap between these two visions isn’t just a communication issue. It’s a ticking clock. It starts with the first funding round and ends with the next forced disclosure, whatever that may be. Founders and serious builders need to keep the two visions as close as possible, or at least close enough to defend if challenged.
What stands out about companies that handle this well isn’t that they’re smarter, they just check in more often. For example, Bezos reviews six-page memos before every big meeting, always focusing on free cash flow. Tobi Lütke at Shopify checks the GSD ledger every month. Frank Slootman at Snowflake ran a weekly internal update called Amp It Up. Brian Chesky manages 50 direct reports, not for fun, but because adding layers between the CEO and customers makes it harder to see what’s really happening.
In successful companies, vision is used to attract talent and customers, while the business model is tracked with just a few key metrics on a regular schedule. Founders who mix up these roles (using vision talk to explain business results) often end up failing.
Stripe is the clearest example. Their goal to ‘increase the GDP of the internet’ isn’t just a vision, it’s their business model. Their developer-first approach shapes how they reach customers. Features like seven lines of code for integration and strong developer documentation aren’t just stories; they’re part of the model. The vision is built into how they distribute their product. Similarly, ChatGPT’s rapid growth to 100 million users happened because they chose an easy-to-access consumer URL instead of a gated API. In both cases, the vision was the channel.
This is what the ‘vision is bad’ story gets wrong. These cases shouldn’t just be warnings, but examples to learn from. Vision alone doesn’t sink companies. It’s when vision isn’t tied to the way the business actually works that problems happen. The real issue is when the story and the business structure aren’t connected.
What Operators See Before the Journal Does
After spending a lot of time inside different companies and markets, I’ve learned that the gap between the two visions is always clear before it goes public. The real question is who has the power to do something about it.
The first thing I check is the ratio of slides in the board deck. If there are 30 slides about vision and only two about unit economics, but most of the discussion is about the numbers, that’s a bad sign. This isn’t just a metaphor. The discussion shows what the board knows needs attention, while the deck shows what founders are willing to share. The gap between them is the real-time disclosure firewall.
The hiring sequence tells you the same thing in a different register. When a Chief Storyteller or VP of Brand is hired before a Head of Pricing or Head of Customer Operations, the cap table is funding the wrong function. Olive AI hired narrative-shapers ahead of implementation rigor and received a KLAS grade that made its positioning untenable. Stripe hired a CFO before scaling its sales organization. These are not accidents of personality or preference. They are structural choices about which discipline gets institutionalized first.
A more subtle but reliable sign is how the language in investor updates changes over time. In companies with a gap between vision and reality, the terms in updates shift every few quarters. MAU turns into DAU, then ‘engaged DAU,’ then ‘qualified engaged DAU.’ ARR becomes ‘strategic ARR,’ then ‘committed ARR.’ These changes are presented as improvements, but really, they’re new metrics. When the definitions change, it means the company is measuring something different, and the gap between the two visions is growing. The language is just being tweaked to hide it.
One of the simplest things a founder can do is look back at past investor updates and compare the language now to what it was 18 months ago. It’s an easy way to spot problems, but hardly anyone does it.
The pattern I find hardest to talk about is the 18-month insider warning, because it involves more than just founders. In every detailed post-mortem, insiders had raised concerns 18 to 48 months before the collapse became public. For example, Ian Gibbons at Theranos pointed out problems years before the WSJ story. EA leaders were warned about Sam Bankman-Fried’s actions from 2018 on, but ignored the warnings. At Olive AI, mid-level operations staff described issues internally that never made it into investor updates.
This warning sign is consistent. Founders often respond by demoting or letting go of the people who raise it. Companies that keep the two visions aligned listen to these signals. Those that don’t usually get rid of the messenger and bring in someone to make the vision sound better.
Where Vision Tolerance Is Actually Structural
After working in many different capital markets, I’ve noticed that saying ‘visionary founders are an American problem’ isn’t actually true, even if it sometimes feels that way.
In 2021, Silicon Valley invested over $300 billion in venture capital. This huge amount of money creates an environment where big stories can last long enough to either succeed or fail in public. In contrast, European pension funds put only a tiny fraction of their assets into venture capital, much less than in the US. In India and Southeast Asia, companies earn much less per user, so the numbers have to work out much sooner. In China, most venture money now goes to state-backed hardware and infrastructure.
This doesn’t mean American founders are more reckless. It just means the US has a unique capital structure that lets big stories last long enough to either work out or fail in public. Other places correct these gaps in different ways. European investors are more cautious, India and Southeast Asia have lower revenue per user, and China uses state controls. The founders aren’t different; the constraints just show up earlier and in different forms.
Klarna faced its correction before its US buy-now-pay-later peers, not because Europe is more disciplined, but because its ownership structure forced a revaluation sooner. BYJU’s in India went from a $22 billion valuation to a rights issue at $225 million, with the founder later saying the company was worth nothing. It’s the same pattern, just a different way and timing for the correction.
The real question for founders today isn’t whether to have a vision, but whether that vision is built into how the product is delivered, or just added later as a story for investors. If it’s just a story, it might help with recruiting, but it’s not a real advantage, and it won’t last forever.
The Clock Is Already Running
What I keep coming back to is that it’s not hard to spot these problems from inside a company. The signs show up months before outsiders notice: the ratio of slides in the deck, the order of hires, changes in language, and the insider who stops speaking up.
The hard part is that spotting these issues means seeing the two visions as a constraint you have to manage at the same time, not one after the other. Founders who say they’ll fix the numbers after they scale aren’t being optimistic. They’re choosing to let the gap grow until they’re forced to reveal it.
Vision isn’t the problem. It’s what attracted the engineers who built Stripe’s simple integration, drove the traffic that made ChatGPT’s launch a success, and helped Netflix keep its content spending steady at $17 billion a year. In all these cases, the vision and the business model were reviewed together, by people who could speak up if they started to drift apart.
The countdown was always there. The only difference was who was paying attention to it.