In 2001, a startup called Webvan shut down after burning through over a billion dollars. Its mission was ahead of its time: on-demand grocery delivery at scale. The team was pedigreed, and the investors piled in. But from launch to collapse, it lasted about 5 years.

Possibly, 4 years too long.

Warehouses were overbuilt, unit economics were misunderstood, and consumer behavior hadn’t caught up.

Years later, people would point to Instacart and say Webvan was simply too early.

But I think that explanation ignores what actually happened. Instacart didn’t own warehouses. It didn’t try to vertically integrate too soon. It piggybacked on existing infrastructure and shifted the model to something the market could support.

The ideas weren’t the same...and that difference is everything.

Webvan failed for one simple reason: because its model wasn’t sound.

By the time that became clear, too much had already been spent on a future that wouldn’t arrive. Too much to admit it. Too much to bail. Too much to walk away, etc.

Founders, especially those still cradling a dying idea, might do well to ask themselves:

If this were going to work, wouldn’t it have worked by now?

You made promises. You built something beautiful, even if no one wanted it. Worse, you convinced other people to believe too - investors, co-founders, a few loyal users.

Admitting it's not working feels like betrayal.

But the clearest signal is the silence. The users who never return, the investors who stop responding, the product launches that yield a few claps on Product Hunt and a polite shrug from the market. And we interpret this as a challenge rather than a verdict. We tinker, we pivot, we relaunch.

There is a time when perseverance is virtue. And a time when it's inertia. Many builders can't tell the difference.

Hope is embedded in the model. "We're pre-revenue." "We're still testing messaging." "We're in stealth." These aren’t red flags, so much as a form of camouflage. Startups are supposed to look broken in the early days. Sometimes it's not the phase. It's the idea.

There's a common genre of postmortem in tech: the noble failure. The idea was great. The team was smart. The vision was right. It was just a matter of timing. A few years later, someone else builds something adjacent and it takes off. And the founder concludes, wistfully, "We were just too early."

Maybe.

Or maybe you were wrong.

Maybe the second team built something different in ways that mattered more. Maybe they didn't believe in your idea, they just believed in your market. It's hard to tell. But the myth of being "too early" is dangerous because it preserves our sense of judgment. It wasn’t the idea. It was the world.

When someone breaks up with you and says, "I just need to work on myself," they may mean it. But more often, they mean they don't want to be with you, and that’s the softest excuse available.

The startup version of that is: "The market wasn’t ready."

But if it were going to work, you would’ve seen signs by now.

Early-stage startups live in the liminal space. Not dead, but not alive. Unresolved, maybe. The team keeps working. There's a little revenue. Some retained users. A roadmap. And no growth.

If you don't look too hard, it could still be a unicorn. And so you don’t look. You avoid honest metrics. You frame downswings as "learning periods." You stretch the burn by slashing marketing, which of course reduces the top of funnel, which you explain as strategic focus.

You get used to lying a little. Mostly to yourself.

In 1519, Hernán Cortés supposedly burned his ships after arriving in Mexico, forcing his men to conquer or die. A brutal allegory for commitment. Founders love that story. But we forget that Cortés already had intelligence suggesting the local civilizations were unstable and politically fragmented. He had a plan. It wasn't just bravado.

Sometimes, burning the ships means you die on a beach for no reason.

Founders tend to measure failure in financial terms. How much equity was wasted. How much investor goodwill was spent. But there's a deeper cost: attention.

While you nurse a non-viable startup, you're not learning new markets. You're not building new conviction. You're not recruiting co-founders for your next thing. You’re burning cycles on something you already suspect won’t work. And worse, you're getting used to it. The abnormal becomes normal, and stagnation is sticky.

Even worse: your taste atrophies. You get used to low uptake, so when something is mildly successful, it feels like a win. You lower your standards. You celebrate being "ramen profitable" long after it matters. You tell yourself the next feature will change everything.

Ideas are not yoga poses.

You don’t get points for holding them longer.

Every investor has a story about the founder who gave up too early. But for every one of those, there are ten who stuck around too long. That's the part we don't talk about.

Markets are vicious. Some ideas are structurally unattractive. Some problems aren’t painful enough. Some customer behaviors are too entrenched. Some business models are just fundamentally f**ked. These facts don’t yield easily to grit - they are not bugs in the simulation, they are the simulation.

An early-stage startup is an information discovery process. The point is to test an idea, not to romance it. But we fall in love, and we mistake signals for fate, and we assume that because something has been hard, it will eventually be worth it. This is the fallacy of sunk suffering.

There are moments in every founder's life where quitting is not failure, so much as allocation. One of the hardest skills in tech is distinguishing between persistence and delusion. And often the best founders are not the ones who refuse to give up, they're the ones who give up with precision.

They know the point where it would’ve worked by now.

You wrote a thousand lines of code. You built a clever backend. You sat through hundreds of user interviews. And yet no one clicks, no one converts, no one stays.

It might feel unfair, and maybe it is. But markets aren’t designed to be fair. They’re not teachers handing out grades for effort. They exist to measure value, nothing more. Your job isn’t to earn a trophy for trying; it’s to discover something people want - urgently and wholeheartedly - and deliver it.

When they don't want what you're building, you can either resent them, or listen. Only one of those leads to better ideas.

It is painful to admit that people don’t want your product. But it is more painful to spend years pretending otherwise.

There is no credit for partial answers.

Maybe your startup isn’t working. Maybe it’s never really worked. That doesn’t make you a bad founder. It makes you a founder who tried something hard and learned something true.

And maybe it’s time to try again. With clearer eyes.

Because if it were going to work, it would have worked by now.