The obvious signals that a team is in trouble — missed deadlines, attrition, escalating incidents — are obvious precisely because they’re lagging indicators. By the time they’re visible, whatever was causing them has been present for a while. You can respond to them, but you’re responding to a problem that’s already been compounding.
The signals I’ve come to pay most attention to are earlier and quieter. They don’t show up in metrics. They don’t surface in project status updates. They surface in skip-level conversations, in the texture of how managers talk about their teams, in what I notice when I drop into a design review. If you’re not looking for them deliberately, it’s easy to walk right past them.
The design review question problem
One of the earliest signals I’ve noticed is a drop in the quality of questions asked in design reviews. Not in frequency — people can ask lots of questions while the quality goes down — but in the kind of thinking behind them.
When a team is in good shape, design review questions tend to be generative: they push on assumptions, they surface edge cases that the designer hadn’t thought of, they ask about tradeoffs between approaches. When a team is under stress or disconnected, the questions become more formulaic: checklist questions, process compliance questions, questions that feel like due diligence rather than genuine engagement. The review happens but nobody is really in it.
I notice this in the design reviews I join across my org, and I hear about it from my managers when they describe their team’s technical conversations. It’s a meaningful signal because it tells you about the team’s collective mental bandwidth. People don’t ask good questions when they’re depleted, when they don’t feel safe enough to push back, or when they’ve stopped caring about the outcomes. Each of those causes has a different fix, but the signal is the same: something is wrong with the energy in the room.
What comes through in skip-levels
I do regular skip-level conversations with engineers across my org — not to circumvent my managers, but to keep a calibrated read on what’s actually happening at the team level. There’s a specific shift I’ve learned to listen for: people move from ownership language to observer language. They stop saying “we’re trying to figure out X” and start saying “I’m not sure what the plan is for X” — stop saying “I think we should do this” and start saying “whatever the team decides.”
This sounds subtle. It is subtle. But it’s a meaningful sign that someone has mentally stepped back from the work — they’re no longer identifying with their team’s outcomes. Sometimes this is about their specific role and can be fixed with a targeted conversation between them and their manager. Sometimes it’s a sign of something broader happening on the team. Either way, I flag it with the relevant manager and follow up explicitly rather than assuming it’s a temporary mood.
I also pay attention to how engineers describe their manager in these conversations — not to gather complaints, but because that language tells you a lot about whether a team has direction and psychological safety.
The absence of informal conversation
The signal I find most reliable, and hardest to describe in a way that sounds rigorous, is the quality of informal interaction. When a team is functioning well, there’s a particular kind of low-stakes conversation that happens constantly — not in scheduled meetings, but in Slack channels, in hallway moments before a meeting starts, in the side comments during a demo. People share small wins, they make jokes about the codebase, they ask quick questions and get quick answers. It’s the connective tissue of a team that trusts each other.
When that goes quiet, something has changed. Not the formal channels — those keep running. But the informal texture changes. Conversations become more transactional. Channels that used to have a lot of noise go unusually quiet. My managers notice this first, and when two or three of them independently mention a similar pattern in their own teams, it’s worth paying real attention.
I can’t always tell you exactly what’s causing it when I notice it — sometimes a specific interpersonal conflict, sometimes accumulated frustration nobody has named, sometimes a few key people quietly burning out and the rest of the team taking its cue from them. But when I notice this pattern and let it pass without inquiry, I almost always regret it.
What I do when I notice these signals
The mistake I made earlier in my career was waiting for the signals to become concrete enough to name a specific problem. By the time I had that certainty, the problem had grown.
What I do now is get curious sooner — not alarmed, not urgent, just paying more attention. I ask my managers direct questions: “I’ve noticed the energy in your team’s design reviews has changed over the last few weeks. What are you seeing?” When something still seems off after that conversation, I’ll schedule skip-levels with a few of the engineers involved — not as an investigation, but as a genuine check-in. The person who seems fine in status updates might be the one already halfway out the door.
Sometimes I’m reading too much into normal variation and nothing is actually wrong — and that’s fine, because a few extra conversations are low cost compared to waiting until missed deadlines make the problem obvious. The organizations that stay healthy longest are the ones where leaders at every level notice something is off before the people experiencing it think to say something. That noticing is most of the job.