You can't install a habit in an afternoon.

Most AI training transfers knowledge in a single session, then expects behaviour to change on its own. It rarely does. Capability is a habit built through repetition on real work, not enthusiasm in a room.

I sat in on an AI rollout for an operations team a few weeks back. The training was good. Genuinely good: a sharp facilitator, real examples pulled from their own work, people leaning in and asking the right questions. Everyone walked out of the room able to explain exactly what the tool did and why it mattered.

I checked back three weeks later. Of the eleven people who had sat in that session, five were still using it. The rest had quietly drifted back to the old way of doing things.

The usual reflex is to blame the training, or the people, or the tool. None of those was the problem here. The session was the best part of the whole exercise. The gap was the bit nobody had planned for, which is everything that happens after the room empties.

Knowing is not doing

A workshop transfers knowledge. In a couple of hours you can teach someone what a tool does, where to click and the sort of jobs it handles well. That part is genuinely easy, and it is where almost all the budget tends to go.

But knowing what a tool does and changing how you actually work are two different things. The first is information. The second is a habit, and you can't install a habit in an afternoon. Habits get built the slow way, through repetition on real tasks, until the new path is simply easier than the old one.

On the Monday after the training, nobody is thinking about the workshop. They are thinking about the forty things in the inbox and the report due at three. Under that kind of pressure people reach for whatever is most familiar, and the familiar option is the old way, every time.

What actually shifts behaviour

The teams that make it stick treat the training as the start line, not the finish. A few things make the difference, and none of them cost much:

• Pick one real task, not the whole job. Choose a single recurring piece of work everyone already does, and run only that one the new way for a fortnight. One habit at a time.

• Give people someone to ask. Most people abandon a new tool at the first small snag. A nominated person who can answer a thirty-second question keeps them moving instead of quietly giving up.

• Make the old way slightly harder. If the old spreadsheet is still sitting there, people will use the spreadsheet. Retire the old path, or at least stop pointing everyone back to it.

• Watch behaviour, not attendance. Whether eleven people turned up tells you nothing. Whether the work is being done the new way a month later tells you everything.

None of this is about motivation. The people in that room were motivated. They were also busy, and busy beats motivated nearly every time. The job of a good rollout is to make the new way the path of least resistance, so that nobody has to feel motivated to take it.

The honest version

If you are about to spend money getting a team trained on something, it is worth asking where that money actually goes. If all of it buys a great session and none of it buys the four weeks that follow, you are paying for enthusiasm. Enthusiasm fades by the following Thursday.

Capability doesn't come from the room. It comes from what you build around the room. Plan that part with the same care you put into choosing the tool, and far more of it survives contact with a normal working week.