You tell engineers and engineering leaders to "just add this one feature, it's really important, what's the ETA? No, that's too late, the market will have moved on, this is existential, you're a bottleneck," fire, hire, fire, hire. Meanwhile the CEO feels great, because they're delivering. The reports look nice, the posts get written, the lectures get given on X and LinkedIn about what steering a ship really means.
Then the issues that required your system to actually be robust, the ones that required engineers to have spent the time engineering, show up. And it's way too late. You cannot move faster, no matter how much money you throw at it.
It gets bad enough that everyone is now telling the CEO 70% of the platform needs refactoring, and even that is like solving the three-body problem, because every domain boundary is broken. Claude can't reason about any component of the system but is very confident it understands it, and ten seconds later tells you "You're right! And I won't sugarcoat it, I completely misrepresented this concept, let me go fix that by rewriting this other module."
You can't rebuild GitHub. You can't rebuild components of it in reasonable time over live data and migrate over without very public incidents. So now the CEO feels very, very dumb, and can't admit it.
Meanwhile a new company capitalizes on your lack of leadership and launches a competitor...and the cycle starts anew.