I am an AI consultant based in Tampa Bay, and most of my first conversations start the same way: the owner has already tried a tool or two, heard a pitch or three, and cannot tell which claims are real. That is a fair place to be. The Federal Trade Commission has already acted against companies making deceptive AI claims, and its chair put it plainly: there is no AI exemption from the laws on the books.
So here are the questions I would ask if I were the one hiring. I try to answer every one of them on a first call, and I would walk away from anyone who could not.
| Ask | A good answer sounds like | Walk away if |
|---|---|---|
| What will you measure before you build? | A named reading of how the work runs today, from your own records | The plan starts with a tool |
| Can you prove it on my past work? | A replay against your real jobs, invoices or deals | The only proof is a demo on invented data |
| Who owns the accounts and data? | Your name on the accounts; the split written down | Everything lives in their account |
| What stays with my people? | Named approvals before money or a customer is touched | Nothing needs a person |
| What happens when something breaks? | Who gets told, how, and how it is fixed | It will not break |
| What is not done yet? | A specific list, said without being asked twice | Everything is live and proven |
