A shared tool is not a shared business

I fed a long, free-flowing brainstorm into my diagnostic setup — the kind of conversation that jumps between half a dozen directions and never quite lands. It came back with a single, confident strategy: one thesis, one funnel, one recommendation. It read well.

It was also wrong. The conversation held about eight distinct ideas, and the tidy version had welded several of them together — ideas with different customers, in different markets, that didn't actually connect. The synthesis looked more finished than the thinking underneath it. Worse, it buried the strongest idea inside a story it didn't belong to, and the headline recommendation came apart the moment you read it twice.

The pull toward one answer is strong, and it isn't only the tool's. It's mine too. A single recommendation feels like a conclusion; eight separate verdicts feel like unfinished work. But the eight ideas were genuinely separate businesses. They happened to share one underlying tool, and a shared tool is not a shared business. The buyers and the markets were different for each.

The fix was dull and correct: pull the ideas apart and give each its own verdict. Some were worth building, most weren't, one was strong alone and weak welded to the rest. No single thread ran through them, and pretending one did had made the whole thing less useful, not more.

I've since changed the tool so a messy, many-idea conversation gets reviewed idea by idea by default, and only gets tied together if I ask for it. The judgment worth keeping is smaller than it sounds: when the neat answer arrives too easily, that's the moment to distrust it.