Capability Beats Access

The licence email went out in February. By March the dashboard showed ninety-four per cent activation, which the leadership team read as adoption and I read as ninety-four per cent of people having clicked a link in an email.

So we spent a Tuesday afternoon watching what people actually did. Not a survey — a chair pulled up beside eleven desks between one and five o'clock. Here is what an ordinary afternoon looked like.

The afternoon

• Four people had not opened the tool that week. Three of them described themselves, unprompted and slightly apologetically, as "not really a tech person."
• Five used it as a better thesaurus. Paste a sentence, ask for a smoother one, paste it back. Real value, roughly the value of a good spellchecker.
• Two were doing something else entirely. One had it summarising every call transcript against the same four questions she cared about. One had built himself a checking habit: draft alone, then ask for the three strongest objections, then rewrite.

Those last two were not more technical than the other nine. I checked, because I expected them to be. What they had was a repeatable move they trusted, and a moment in their week where it belonged.

Access is a purchase. Capability is a practice. The invoice cannot tell the difference, which is why the invoice keeps getting paid.

What actually moved the number

We stopped running training. Training was the thing that had already failed twice: a ninety-minute session, a slide about hallucinations, a demo of a feature nobody needed that day. Instead we asked each team for one task — one, named, recurring, ideally the one that everybody groaned about — and we built the habit for that task only, in the meeting where it already happened.

For the client services team it was the Monday summary. For the engineering leads it was reviewing a design doc against a checklist they had written themselves. Eight weeks later both teams had spread it sideways on their own, which is the only kind of spreading that has ever held.

If you are trying to decide what to spend the next quarter on, I would not spend it on more access. Everyone has access. Spend it on one Tuesday afternoon of honest watching, and then on the single task that afternoon makes obvious.