When the playbook stops working

The interesting problems are the ones that don't give way to the moves that have always worked for you. Most operators meet a few in a career, and they're disorienting in a specific way.

Here is the shape it tends to take. A capable operator with a long record of fixing things hits something that won't budge. They reach for the obvious cause and throw the usual fix at it, hard, the way that has always worked. Nothing. So they try the next obvious thing, then the next, often with their best people on it, and somewhere in there it stops feeling like problem-solving and starts feeling like flailing. For the first time in a long while, they are trying everything and missing, and can't say why.

I have watched this from close up. A general manager with flat growth pushes sales, because pushing sales is what you do, and the team posts record numbers, and the business still doesn't move. The reason is upstream and unglamorous: customers were being oversold at the point of sale, so the growth was leaking out the back as fast as it arrived at the front. No amount of better selling was ever going to fix a problem that selling was causing.

The thing worth noticing is that the operator was not wrong to be good at their repertoire. The repertoire is real, and it earned them the seat. The problem was simply a different kind than the ones it was built for: judgment-heavy, badly structured, the sort that has no known move you can execute well. You can be excellent at the kind of problem that rewards a practised response and helpless in front of the kind that first rewards working out what kind of problem it even is.

So when something resists everything you throw at it, the useful move is usually not another, harder go. It's to stop and ask whether this is the kind of problem your track record was built on at all. The ones that resist are almost always resisting for that reason.

The skill that built your career can be the thing standing in front of the problem that won't move. When the playbook stops working, the problem hasn't got harder. It's got different.