Start with the initiative, not the model

I have watched the same story play out many times. A team buys an AI tool, runs a pilot, and the demo looks great. Six months later nobody can say what changed. The tool usually was not the problem. It just was not attached to anything the business was measuring.

The fix is to reverse the order. Start with an initiative the executive team has already committed to, then work backward through outcomes, process and data until you reach the technology. By the time you pick a tool, you know exactly what it has to do and how you will judge it.

1. Name the executive initiative

Pick something leadership has already committed to, ideally in writing: the annual plan, the board deck, this year's goals. If it is not on that list, it will not get budget, attention or people's time when things get hard.

Good initiatives sound like "grow recurring revenue 20 percent," "open a second location without adding back-office staff," or "pass our first security audit by Q3." A bad one sounds like "use more AI."

2. Turn it into outcomes you can measure

Write down one to three numbers that would prove the initiative is working, with today's baseline, a target and a date. If you cannot find a baseline, measuring it is your first project. Skipping this step is how pilots end in opinions instead of decisions.

3. Map the process and capabilities

Walk the work as it actually happens, not as the process document says it does. You are looking for the places technology can move one of your numbers:

• Where does time pile up, and who is waiting on whom?
• Which steps involve someone reading, summarizing, sorting or re-typing information?
• Which decisions stall because the information arrives late or incomplete?
• Where do errors get caught, and how much rework follows?

4. Find the data and systems

For each step that matters, list where the information lives and who owns it. This is where most projects are won or lost. An AI assistant that cannot see the customer record, or cannot write its result back, is a demo, not a solution.

5. Choose the technology last

Now ask what the smallest set of tools is that closes the gap. Sometimes that is an AI model. Often it is an integration, a better form, a report, or switching on a feature in software you already pay for. AI earns its place where the work involves reading, writing, classifying or deciding on unstructured information at volume.