Where to start with AI — a $1,500 first step.

The first AI decision most businesses get wrong is scale. They start too big, spend too long, and learn nothing until the money is gone. Start small enough to be wrong cheaply.

Pick the workflow by these four tests

• High volume. It happens dozens of times a week, not twice a month. Volume is what makes automation worth the setup.
• Rule-based. The right answer already exists somewhere in writing. If your team argues about the correct answer, AI will not settle it.
• Low blast radius. If it gets one wrong, someone is mildly inconvenienced — not harmed, not sued.
• Measurable. You can count it today. If you cannot count it, you will never prove the project worked.

Answering repeat customer questions passes all four in almost every business. That is why it is the standard first build.

What the first two weeks look like

Week one is scope: naming the workflow, gathering the content it answers from, and agreeing what “good enough” means as a number. Week two is the build and the first real conversations, watched by a human who can step in.

By day fourteen you have something running and a number to argue about. Compare that with a strategy engagement, which at day fourteen has a calendar invite.

What to do with the result

If it works, you have earned the right to do something bigger — and you now know your content, your team and your customers well enough to scope it properly. If it does not work, you have learned that for $1,500 rather than $150,000, and the reason will be specific enough to fix.