If you read the death-flailing post and thought "yeah, that's us," this one is for you. Those early signs - the eerie quiet, the rotating saviors, the vanity metrics - show up when a company still has a shot at course-correcting. The signs below tend to show up later, when the window is getting smaller and people are starting to do the math on their own.
I've been the programmer inside startups since 1999, and the pattern doesn't change much: by the time someone says any of this out loud, it's usually late.
The roadmap only moves right
Every quarter, the big features slide to the next quarter. Not because the team is slow, and not because priorities genuinely changed. They slide because the roadmap exists to manage investor optics, not to ship software.
You can tell the difference. A real reprioritization comes with a reason someone can defend in a room. A slide-to-the-right comes with a revised deck and silence, because asking "why" out loud is how you end up on the next layoff list.
When the roadmap is fiction everyone pretends to believe, the product stops mattering. The story becomes the product. You're not building software anymore - you're writing a pitch deck that updates quarterly.
Recoverable? Sometimes, if someone with authority is willing to burn the roadmap and rebuild it from customer conversations. Rarely happens without a leadership change.
Headcount is growing faster than revenue
Hiring is optimism made visible. I get it. You close a round, you post jobs, you staff up for the future you believe in.
The problem is when headcount keeps climbing and revenue isn't keeping pace. Not falling behind slightly - I mean the ratio is getting worse every quarter. More people, same revenue, or worse, less revenue per head.
What this usually means in practice: the company is buying time. Each new hire is a signal to investors that things are moving, even when the underlying numbers aren't. It's activity theater. And it works, right up until the moment the next round doesn't close.
The people hired into this are the ones laid off six months later in the "restructuring." The tell is the big morale event right before the cut: the offsite, the all-hands, the new vision deck. That isn't bad timing. It buys one more week of nobody asking questions.
Recoverable? Only if leadership freezes hiring and gets honest about the unit economics before their runway runs out. Most don't.
The founder starts talking about "the narrative"
Founders need to tell a story to investors and press. That's fine.
The version that isn't fine is when "the narrative" starts showing up in internal conversations. When the founder is more focused on how the company is being perceived than on what the company is actually doing. When the answer to "why aren't customers converting?" is "we need to tell a better story" instead of "we need to talk to more customers."
You can often see it from the outside. The founder's LinkedIn posts are getting thousands of likes, the customers are churning, and nobody at the top seems to see the connection.
Founders who get lost in the narrative aren't lying, exactly. They've started believing their own pitch deck. They think if the story is good enough, the product will catch up. It won't.
Recoverable? Rarely, without the founder getting very uncomfortable feedback from someone they actually respect. That person is hard to find when you're surrounded by people whose jobs depend on agreeing with you.
Customer calls get replaced by dashboards
Early on, founders talk to customers constantly. It's annoying and time-consuming and absolutely necessary.
At some point, the company gets big enough to have dashboards. Retention metrics, NPS scores, support ticket volume, churn rate. All useful. None of them replace a thirty-minute call with a customer who's about to leave.
The shift happens gradually. Customer calls become "customer success's job." The founder stops doing them. Then the exec team stops. Eventually nobody who makes product decisions talks to customers at all. The dashboards get bigger and more colorful and everyone stares at them in the weekly meeting and nobody can explain why churn is up.
Dashboards tell you what happened. Customers tell you why. When you lose the why, you start guessing, and a team can burn six months building the wrong thing because nobody wanted to admit they were guessing.
Recoverable? Yes, and faster than most of the others on this list. Pick up the phone. Call five customers who churned in the last ninety days. Listen without defending. You'll know what to do.
The people who actually shipped things start leaving
This one is the clearest signal on the list, and the one leadership ignores hardest.
Every company has a small group who actually ship. They know which parts of the codebase are held together with prayer. They know which customer is one bad bug away from leaving. They know the CEO's quarterly plan is fantasy.
When those people start leaving, pay attention to why. Not the official reason. The real reason.
They leave because they've stopped believing. They watched the same three features slide four quarters in a row. They sat in planning meetings where the CEO promised a customer something engineering can't build. They've done the math on runway and hiring burn rate. They know how this ends.
The people who leave last are the ones who can't leave yet - they need the vesting cliff, or they've got a mortgage and the job market is rough. When the people with options bail first, the company is telling you something about its own future.
Recoverable? Depends on how many have already gone. One senior person leaving is noise. Three in two months is a pattern. By the time you're offering retention bonuses, you've already lost - you're paying people to stay long enough to finish the handoff docs.
None of these signs are subtle. The hard part isn't seeing them. It's admitting you're seeing them in a company you're invested in - financially, emotionally, or both.
The earlier post covers the early signs - when you might still turn it around. This one is for when you're past that. Some of these you can fix with real changes and uncomfortable conversations. Some of them mean it's over and you're deciding how long to pretend it isn't.
If you're a founder reading this and nodded at more than two, stop reading and start calling customers. Not the happy ones. The ones who churned last quarter.
-Sethers