Most AI marketing consultants sell one of two things: a strategy deck, or access to their own tool. Neither leaves you with a system your team runs on its own. The seven questions below sort the consultants who build for you from the ones who keep you renting.
If you are still working out what the role even covers, start with what an AI marketing consultant does. This guide assumes you have decided you want help and need to pick who.
1. Will you build in my stack, or move me onto yours?
A consultant who moves you onto their platform has built themselves a recurring fee, not a system for you. Ask what happens to your workflows if you stop paying them. The answer you want: everything runs in tools you already own or would choose anyway, under your accounts.
2. Do I leave each session with something running?
Recommendations are cheap. A working workflow is not. Ask what you will have at the end of the first engagement that you did not have before: a content pipeline that drafts in your voice, a reporting loop that pulls your real numbers, an agent that handles a task your team does by hand today. If the answer is "a roadmap," you are buying homework.
3. Who owns what you build?
Get this in writing. The prompts, the instructions, the files, the automations and the accounts should all be yours. If a consultant builds inside their own workspace and hands you a login, you own nothing the day the contract ends.
4. How do you handle our brand context?
This is the question most buyers skip, and it decides whether the output sounds like you. AI that starts every task from a blank prompt produces generic copy. Ask how the consultant captures your positioning, audience and voice so every workflow reads from it. Good answers describe a persistent source of truth the AI reads automatically. Weak answers describe a long prompt someone pastes in each time. We wrote up why that breaks in the brand voice file trap.
5. Can you show me something you built, live?
Screenshots and case-study PDFs are easy to produce. A screen-share of a real system doing real work is not. Ask to watch one run. A practitioner can open their own setup and walk you through it in five minutes.
6. How do you scope and price the work?
Scope should come before price. A consultant who quotes a monthly retainer before understanding your stack is pricing their time, not your problem. Ask how they define done for the first project, and what is out of scope. Our guide to scoping AI projects covers what a sound scope looks like.
7. What happens when you leave?
The best engagement makes itself unnecessary. Ask what your team will be able to change, extend and fix without the consultant: documentation, a handoff session, the skills to add the next workflow on their own. If the honest answer is "call me," you are buying dependency.
Red flags
- A proprietary platform you must subscribe to. Lock-in dressed as convenience.
- No live demo. If they cannot show a system running, they may not have one.
- Retainer before scope. Pricing time instead of outcomes.
- Vague ownership terms. Anything built in their accounts is theirs.
- Tool lists instead of workflows. Knowing twenty AI tools is not the same as building one system that works.
When you do not need a consultant
If you have the time and a specific, bounded goal, you can build a lot yourself. Claude Code reads your files, keeps your context and runs multi-step work, and you do not need to code to use it. Claude Code for marketing is the place to start. A consultant earns their fee when you want it built faster, want a second set of eyes on the design, or want your team trained while it gets built.
How I work
I run 1:1 AI marketing advisory as live screen-share sessions. We work on your actual stack with Claude Code, you leave with something running, and everything we build is yours. Every engagement starts with a free 15-minute intro call where we figure out whether it is a fit and scope the work before any price is set.
Want a second set of eyes on your AI marketing build? Book a free 15-minute intro call →