AI strategy

No, nothing is built at this stage. The result is an ordered list answering the question of where to start, with a budget beside each line. A company that already knows its use case, say sorting the orders that arrive by email, is better off going straight to a pilot instead of paying for analysis.

The heads of the departments concerned, whoever runs your IT, and a member of management who can decide. Having the DPO in one of the workshops helps. If the business side withdraws, what remains is a technical exercise with no connection to real work, and the plan will never be applied.

Schemes linked to France 2030 or supported through Bpifrance periodically cover business digitisation, and a written strategy sometimes helps put a file together. We are not a grant broker and we promise no outcome. The plan is built to stand on its own even if no outside funding arrives.

That conclusion has value: it saves you from funding a project with no return. We would rather write that it is too early, list the preparatory work that would change the picture, such as sorting out a product catalogue, and let you decide. An honest strategy also says what not to do this year.

The tool market moves quickly; your processes and your data move much more slowly. We suggest a short review twice a year: updated costs, new capabilities, regulatory changes. The priority list itself usually lasts longer than people expect, because it rests on how your company works rather than on the fashion of the moment.