why do startups fail with no market need? a sequoia company with one possible customer

today i taped with an operator turned angel who has seen more startups than almost anyone i know. he founded six of his own. four wins, two failures. across his career he's invested in more than three hundred companies, and he's still active in two angel groups in the bay area.

i asked him for the failure that taught him the most. he went back to 1989.

he was at a chip company. sequoia backed. they built chipsets compatible with sun's sparc architecture. good engineering. the product worked.

and then the market didn't show up.

"you either were able to sell to sun microsystems, and if you didn't get that account, you were dead. so sun didn't buy from us and we're actually sequoia capital-backed company."

that's it. that's the whole story.

a viable product. the best venture firm in the world on the cap table. and exactly one possible customer. when that customer said no, there was nobody else to call.

founders ask why do startups fail with no market need, and they picture a product nobody wants. a bad idea. a solution looking for a problem.

his story is scarier than that. the product was wanted. by one company. that's not a market. that's a single point of failure with a logo on it.

i see versions of this all the time. the startup built for one enterprise pilot. the platform that only works if one big partner integrates. the deal that's ninety percent closed for nine months. the founder isn't lying when they say there's demand. there is. it's just concentrated in a place they don't control.

he's distilled what he learned into three tests for deep tech, and they work for any startup.

first, is the founding team commercially oriented? not just brilliant engineers. people who think about who pays and why.

second, is there a real market, and will the timing still be right when you launch? the sparc compatible market looked real when they started. it never took off. you can be right about the technology and wrong about the clock.

third, is there a credible path from technology to product? a lot of great science never becomes something a customer can buy.

he had a red flag i laughed at, because i've seen it too. founders claiming six exits by thirty. when the story sounds too good, it usually is.

he's also wary of where valuations have gone. seed safes with caps in the thirty to forty million range. for him, a five to ten times return is a satisfying outcome for an angel. when the entry price is that high, even a good company struggles to deliver it.

what i took from him is a simple exercise. before you count your features, count your buyers. real ones. companies that could sign, that have budget, that have a reason to choose you this year. if the list fits on one hand, your next job isn't the product. it's the market.

and if your list is short, that's not a death sentence. it's a to-do list. talk to the second and third possible buyers now, while the first one is still deciding. find out what they would need to say yes. sometimes the product barely changes and the market triples. sometimes you learn the market was never there, and it's far better to learn that in year one than in year four.

long term, the founders who survive aren't the ones with the best tech. they're the ones who made sure more than one door could open.

the founder's fundraising checklist makes you answer the buyer question before an investor asks it for you.

so write the list right now... how many companies could buy from you this year, and what happens if the biggest one says no?