Start with the work, not the tool

When a team begins exploring AI, the most natural question is often: Which tool should we use?

It is a reasonable question. But it can pull attention away from the harder and more useful one: What part of our work deserves a better way of doing it?

Starting with the work changes the conversation. Instead of asking people to imagine an abstract future, it gives them a familiar task to examine together. A briefing that takes too long to prepare. A handoff that loses context. A weekly update that requires copying information from three places.

Look for a useful point of friction

The best first experiments are usually small enough to understand and important enough to matter. They have a clear beginning, a recognizable output, and someone who can say whether the result is actually helpful.

That makes learning more honest. People can see what the system does well, where it needs review, and where it does not belong at all.

Make practice social

Confidence grows faster when people can compare notes. Give a team time to try an approach, discuss what happened, and adjust the next attempt together.

This does not need to be a large programme. One shared workflow, a short working session, and a clear record of what was learned can be enough to create momentum.

The goal is not to make every task feel automated. The goal is to help people spend more attention on the parts of work that need their judgment.

Keep the next step visible

A good experiment should leave behind something useful: a prompt that has been tested, a documented decision, a simple template, or a clearer understanding of the workflow itself.

When teams start with the work, technology becomes easier to assess. It is no longer a promise in the abstract. It is one practical tool in service of people doing meaningful work.