Do You Need an AI Consultant? An Honest Test

I sell this, so I'll start with the answer that costs me money: most small businesses don't need an AI consultant yet.

When you don't need one

You haven't used an AI assistant on a real task yet. Spend a month of your own attention before you buy someone else's. Nothing a consultant tells you will stick until you've tried it once and done it badly.

You want AI because it came up in a meeting. That isn't a problem to solve, and a consultant will happily take money to make it look solved.

Your process only exists in your head. Automate a process nobody wrote down and you've built a faster way to be confused.

When you do

You've done the basics and stalled. You can draft an email in thirty seconds and still lose two hours a day. That gap is usually wiring, not effort, and wiring is someone else's strength.

You're paying for tools you don't use. Most owners are. A short audit finds them faster than you will.

The work only happens when you're in the room. That isn't a system, it's a job you gave yourself.

What a good one actually does

Not a slide deck. Not a roadmap you'll never open again.

They watch how you work. The most valuable thing in this business is someone asking "what just happened there?" while you do an ordinary Tuesday.

They write it down. Your steps, your rules, your exceptions. The document matters more than the tool.

They leave you able to change it. If you can't adjust it next month without calling them, you didn't buy a system. You bought a dependency.

What it should cost

For a small business, a first engagement should be small and specific: one workflow, one document, one thing that still runs after they're gone. If the number is large and the scope says "AI strategy", walk.

I start at $750, and the smallest thing I take on is one workflow set up properly. Anything bigger should come after you've watched the first one work.

A cheaper way to find out

Before you hire anyone, do this. Write down the one task you'd pay to never do again. Then spend one evening trying to get an AI to do most of it.

If it goes well, you now know what to ask a consultant to scale. If it goes badly, you now know exactly what to hand over and what it's worth.

Either way you arrive with a real question instead of a budget, and that changes the conversation completely.

If you just want somewhere to start

The Prompt Playbook is free, and the community is free. Both are places to get an answer tonight before you spend anything on advice.