For CFOs

We announced an AI budget. What return should we hold it to?

Hold each piece of it to one thing it replaces, adds or protects, with a euro figure and a date. A programme has no return. The single tools, pilots and hires inside it do, or they do not.

The budget was probably approved the usual way: a total, a list of initiatives, and a sentence about staying competitive. That is a fair reason to start. It is not something you can check in twelve months. The return line can still be written now, one item at a time.

If you had to name the one item in the programme with a number next to it today, which would it be?

Run four questions on one AI project

Two-minute video and a check you can fill in with rough numbers.

Buying a tool, not building?

Three questions before the purchase is signed.

What to look at first

01

A before number for each item

What does the work cost today, in hours and euros, before the AI touches it? Without that, any result next year is an opinion.

02

One owner who signs the return

Not the person who built it. The person whose budget should get smaller or whose revenue should grow. If nobody will sign, that tells you something.

03

A date to stop

Agree now when each item is reviewed and what result means it ends. Stopping is easier to agree before anyone is attached to the work.

The longer version: How do I know if our data and AI spend is paying off?