The Two Causes of Startup Death
A two-item diagnostic: modern startups almost never die of broken technology, they die of founder implosion or of solving a problem nobody cares about.
- Difficulty
- Easy
- Time to result
- ~weeks to results
- Steps
- 5
- Confidence
- 85%
English has been investing alongside operating for a long time and says the failure pattern has changed. In the old days companies failed because the product did not work and they simply could not get the tech working. That is hardly the case now: between AWS, modern tooling, and open source, it is fairly easy to assemble the right libraries and the right people to put something together. So he reduces failure to two causes. One is founder implosion, where the founders create a toxic culture, people hate their boss and hate working there, and everybody leaves. He calls this failing the culture code and says it happens all too often. Two is building a nice product for a real problem that nobody cares about, which he specifically links to the founder who has an itch, gets obsessed, works on it for ten years, and never notices how few people share it. He says it makes him sad to see a beautiful product that solves a problem he does not care about.
Origin
English formed it across roughly 60 angel investments plus six of his own companies, watching which failure mode actually showed up as tooling improved.
Core principles
- 01Technical execution risk is mostly a solved problem now
- 02Culture failure kills companies through the exit door, not the roadmap
- 03A real problem and a problem people care about are not the same thing
- 04Founder obsession is a strength that becomes the second failure mode
- 05Beautiful products die routinely for lack of anyone who cares
How to run it
- 1
Retire technical risk as your primary worry
Accept that the stack, AWS, modern tooling, and open source have made assembling a working product tractable. English's point is that time spent agonizing over whether the tech can be built is time not spent on the two risks that actually kill you.
Watch out This does not hold for genuinely novel deep tech, where the old failure mode is still live.
- 2
Run the founder implosion check
Look directly at whether the founders are generating a toxic culture. The observable symptom English names is people hating their boss, hating working there, and leaving, which he calls failing the culture code.
- 3
Run the nobody-cares check
Separate the question of whether the problem is real from whether anyone cares about it. English's failure case is a nice product built for a real problem that no one cares about, which passes a naive problem-validation test.
- 4
Audit your own obsession
Ask specifically whether you are the founder with the little itch who will work on it for ten years. English says these founders get really obsessed and do not realize that not a lot of people have that same problem.
Pro tip The tell is that all your evidence for demand is your own experience of the problem.
- 5
Fix the failing check before improving the product
If either check fails, product work is the wrong response. Culture failure needs leadership intervention and demand failure needs a different problem, and shipping more features addresses neither.
In the wild
English describes the pattern he sees repeatedly across his investments: a founder has a little problem, gets really obsessed, works on it for ten years, and never registers that not many other people have that problem. He says it makes him sad when he sees a beautiful product that solves a problem he does not care about.
→ A polished, working product with no market, which English names as one of the two dominant causes of startup death today.
Common mistakes
Spending diligence on technical feasibility
Investors and founders default to interrogating whether the team can build it, which English says is hardly ever the binding constraint now. It is a comfortable question because it has a crisp answer, and it leaves both real risks unexamined.
Treating a real problem as proof of a market
The failure mode is specifically a real problem that no one cares about. Validating that the problem exists is not the same as validating that a meaningful number of people would change their behavior or pay to remove it.
Reading founder obsession as conviction
Obsession is what keeps a founder going for ten years, which is exactly why it is dangerous when pointed at a problem nobody shares. Without an external demand check it insulates the founder from the signal that would otherwise redirect them.
Is it for you?
Best for
Founders and early investors deciding what to worry about, and where to spend diligence time
Not ideal for
Deep-tech teams whose core technical risk genuinely is unsolved, where the old failure mode still applies
From the transcript
“in the old days companies would fail because their product didn't work. They just couldn't get the tech working. That's hardly the case now.”
“Companies fail today for one of two reasons. Either there's a founder implosion and they create a toxic culture and people like hate their boss…”
“Or number two, they built a nice product, but it's a real problem that no one cares about.”
“a lot of founders have an itch to scratch as you were saying earlier, Sam. And they get really obsessed about it. Like, I have…”
From the episode
Paul English: How He Built And Sold 6 Companies And Why He Believes Irritation Becomes Inspiration
Paul English