Don't Team for the Plan

There's a management anti-pattern so common it barely registers anymore. Leadership lands on a new strategy — AI, platform consolidation, whatever the moment demands — and instead of asking "do we have the right people for this?", they ask "how do we fit our current people into this?"

That's teaming for the plan. And it quietly destroys teams.

⚔️ The Raid Analogy Nobody Wants to Hear

If you've ever played World of Warcraft, you already understand team composition better than most engineering managers.

Before a raid, you check your roster. You need tanks, healers, DPS — specific specs, specific levels. If you're short a tank, you don't tell your holy paladin to just figure it out. You reschedule. You find a different raid. You adapt the plan to the team, not the team to the plan.

You don't force a heal-specced paladin to tank. They'll try. They'll fail. And they'll know exactly why — even if you won't say it out loud.

Engineering orgs do this constantly. They build a strategy, then look at their roster and start squinting. "You're close enough, right? You're adaptable. Everyone's a generalist here."

That last line is the tell. When leadership suddenly declares that everyone is a generalist, it usually means nobody is — and the team composition is too inconvenient to fix.

🔀 Hybrids Work in a Pinch

This isn't an argument against versatility. Hybrids have their place. You ship with the team you have sometimes, and good engineers can stretch.

But stretching has a ceiling. And when you stretch someone past it for long enough, two things happen:

1. The work suffers — because specialists exist for a reason
2. The person suffers — because they can feel the mismatch on every pull

The issue isn't the hybrid. The issue is treating a pinch solution like a long-term strategy.

🤖 AI Won't Fix a Misaligned Team

Here's where it gets timely.

A lot of orgs right now are betting that AI will paper over their structural problems. If we just add enough automation, enough tooling, enough throughput — maybe it won't matter that we have the wrong people doing the wrong things.

It will matter.

AI is an amplifier. It makes fast things faster. It doesn't fix misalignment — it scales it. If your director is using five minutes between meetings to brain-dump into an LLM and ship a clean document to the team, that's not efficiency. That's throughput moving upstream while the cognitive cost redistributes downstream. Everyone else now has to read, parse, and align on that document. You didn't save time. You borrowed it from everyone else. ⏳

The bottleneck was never document quality. It was attention and alignment. You can't automate your way out of a strategy problem.

✅ What Actually Works

Team makeup should reflect team goals. Full stop.

If goals change, update the team. If you can't update the team, update the goals to match what the team can actually do well. If you can't do either, be honest about the gap instead of renaming it.

That's not pessimism. That's resource planning. It's what any good raid leader does before the pull. 🧠

Plan for the team. Don't team for the plan.