When I joined AWS to lead the EC2 Pricing Engineering team, there were eight engineers. Two years later, there were fifty-plus. That growth sounds clean on a slide deck. Living through it was something else. You don’t move linearly through the Tuckman stages — you loop back through all of them every time you hire a new cohort, every time a reorg reshuffles the team, every time a leadership change sends people back to square one. I’ve lived this cycle more times than I can count. Here’s what each stage actually looked like from inside it.
Forming: The Danger of Too Much Politeness
The team was deep in Forming when I arrived — too polite, too agreeable. Meetings produced consensus without real debate — everyone nodding, nobody pushing back hard on anything. Code reviews were gentle to the point of useless. People deferred to each other in ways that felt collegial but were actually just friction-avoidance dressed up as respect.
Forming feels stable, but that’s the trap. Teams that stay in it too long start mistaking politeness for alignment. They’re not building shared understanding. They’re avoiding the discomfort where the real thinking happens.
My approach was to disrupt it deliberately: I started asking hard questions in design reviews — not because I was sure I was right, but to see how the team handled pushback. I set a norm explicitly and early: “Argue with me when I’m wrong. If you don’t, you’re not helping me make better decisions.” That norm alone usually starts the move toward Storming — which is exactly where you want to go.
Storming: It Doesn’t Always Look Like Conflict
Storming on engineering teams is rarely loud — it’s usually quiet. A senior engineer who stops speaking up in planning sessions because “no one listens anyway.” Two tech leads with competing architectural proposals lobbying me separately instead of talking to each other. A new hire from a different engineering culture, visibly frustrated, processing it alone instead of raising it directly.
The instinct when you notice this is to smooth it over — call a meeting, make a call, move on — but that’s almost always wrong. Smoothing Storming doesn’t resolve it; it drives the tension underground, where it becomes passive resistance and, eventually, attrition.
What actually works is naming the tension out loud. “I’ve noticed we have two architectural directions and the team is split. We’re going to evaluate both properly. I want real debate — not backroom lobbying.” Giving the conflict a legitimate forum is how you move through it, not around it.
Norming: Invisible Until It’s Gone
You know you’ve hit Norming when people stop asking “how do we do this?” and just do it. Design docs follow a consistent structure nobody explicitly mandated. On-call rotates and organizes itself. New engineers get ramped up by the team — not by waiting for you to have a free hour.
Norming is quiet and productive, which makes it dangerously easy to take for granted.
The mistake I see leaders make in this stage: going hands-off at exactly the wrong moment. Norms live in people’s heads, not in documents. Every significant hiring wave is a partial reset — the new people weren’t there when those norms formed, and they don’t absorb them by proximity. Now, after any meaningful growth event, I explicitly re-establish working agreements. We revisit how we define done, how we handle technical debt, how we escalate blockers. It takes overhead I sometimes resent spending, but it’s consistently worth it.
Performing: The Goal, and Why It Doesn’t Last
At peak, EC2 Pricing Engineering was shipping faster than any team I’d seen at that scale. Engineers were making high-quality architectural decisions independently, getting them peer reviewed by other senior engineers — no manager in the loop as a bottleneck or tie-breaker. Incidents got resolved before I knew they’d started. Cross-org negotiations happened without routing through me.
Performing is intoxicating, but it’s also impermanent. Every reorg, every leadership change, every major strategic shift pushes you back — sometimes to Forming, sometimes straight to Storming if the change is contentious enough.
Performing isn’t a destination you reach and keep — it’s a condition you maintain by actively watching for the early signs that you’ve already started to slide. If you’re not scanning for Storming signals when things feel healthy, you won’t catch the regression until you’re already deep in it.
Tuckman is not a ladder — it’s a cycle. The leaders who get the most out of it are the ones who can read where their team actually is — not where the org chart says they should be. That requires honesty and more watching than talking.