The best opportunities tend to be problems hiding in plain sight. Three books explain, from different angles, why that keeps happening.
The Stripe insight from How Stripe Got Started: payments weren't solved, they were just tolerated. This was 2010. PayPal existed and was large and growing, but as a developer platform it was neglected. The alternatives were legacy bank relationships and painful APIs that required weeks of integration work. Developers had accepted this as the cost of doing business because it had always been that way and PayPal was "good enough." What makes that even sharper is that a thriving, profitable incumbent left the integration pain fully in place because it didn't have to care. When something has been painful long enough, the pain stops registering. It gets reclassified as just how things work.
If something is universally annoying and nobody has fixed it, that's worth looking at. Low expectations are a form of camouflage.
The Innovator's Dilemma explains why incumbents miss disruptive opportunities even when the threat is visible. Disruptive tech starts worse on every metric that matters to existing customers — lower margin, simpler, aimed at a market too small to move the needle. Rational resource allocation keeps pointing away from it. The companies that should solve the problem are structurally prevented from doing so.
The same thing happens at the individual level in engineering organizations, and it's more cynical. An implementation is in place with significant known problems. Everyone knows what the fix is. Nobody does it — not because the problem is invisible, but because leading a big refactor is a career risk. If it goes well, the benefit is diffuse. If it goes sideways, you're the person who broke things. Easier to work around it, stay heads-down, and collect a paycheck.
It looks like a knowledge problem but it isn't. The information is fully public and the fix is agreed upon. What's missing is someone willing to absorb personal downside for a collective benefit, and that's an incentives problem.
The Wide Lens adds a third failure mode: even when someone builds the right solution, the ecosystem around the problem blocks it. Your success depends on other innovations also succeeding, and every intermediary between you and the end customer has to say yes. One weak link breaks the whole chain. The product was fine. The problem was everything around it.
Each failure mode is also an entry point. If a problem has been reclassified as normal, you're one of the few still seeing it as a problem, which means the camouflage is working in your favor. If the people best positioned to fix it are incentivized away, the field is empty for whoever is willing to eat the downside. And if the ecosystem isn't coordinated, that coordination becomes the moat once you're the one who does it.
None of this is easy, which is the point. The obviousness of the problem tells you nothing about the difficulty of the solution, and that gap between "everyone can see it" and "almost no one will do it" is where the opportunity lives.