What Does an AI Marketing Consultant Do?
By Adam Sandler, founder of The Viable Edge · Updated
An AI marketing consultant maps how your marketing team works, chooses where AI earns its place, builds those workflows on your own sources and tools, and trains your people to run and review them. In my practice that is five steps: Diagnose, Ground, Build, Train, Review. People stay in the decisions at every step, and the aim is a team that does not need me ongoing.
What does an AI marketing consultant do, step by step?
The Viable Edge’s AI-Native Marketing Team Framework is the order I work in. Each step has a job, and the first one always comes first.
- 1
Diagnose
Map the work and choose where AI earns its place. It always comes first, and it is lighter for a small team.
- 2
Ground
Bring together the context the work depends on: brand and business context, sources, examples and standards. A knowledge base can help; it is not required.
- 3
Build
Wire the workflow to the tools your team already uses, and test it on real work.
- 4
Train
Train on real work, per build, so every system ships with a trained owner on your team.
- 5
Review
Check what is used and what changed. You decide what to expand.
Diagnose takes the first two weeks: a review of the tools and workflows your team already uses, then a prioritized plan for where to start.
The workflow I build first most often is brief to draft to review. Content or campaign briefs become drafts, and the drafts go through review.
Grounding comes before building, so the workflow draws on your team’s real sources instead of output that only sounds right. A named person on your team owns that source material.
What do you get at the end?
You end with workflows your team runs, the sources they draw on, the review points you agreed, and people trained to own them.
- A working workflow for one agreed piece of recurring work, wired to the tools your team already uses.
- The context and sources it draws on, written down.
- The approval steps your team agreed, built in.
- A named owner on your team, trained on real work to run it.
- A review against the baseline we set at the start, with you deciding what to expand.
Here is the shape of one client workflow, with the details anonymized.
A client workflow brings organic performance reporting and a shared content store into one process. Data and business context inform a brief; drafts carry that context forward into human review. The work has clear sources, and the team can resolve gaps before preparing content for publishing.
- 1. Performance data
- 2. Source-backed brief
- 3. Context-informed draft
- 4. Human review
Example drawn from client delivery. Details are simplified and anonymized. This illustrates workflow capabilities, not a claim about revenue or time saved.
Where do people stay in the decisions?
People stay in the decisions in four specific ways, agreed before anything is built.
- Nothing goes out to a customer or the public without a person approving it.
- A named person owns the source material the AI works from.
- People review at set checkpoints (brief, draft and final), not only at the end.
- Strategy, positioning and priorities are decided by people. AI drafts and assembles.
Anything that would publish, send or spend waits for a named person on your team to approve it.
How is an AI marketing consultant different from an agency or a software vendor?
The difference is who does the work and what you keep. An agency runs the work for you. A software vendor sells you a platform. A consultant builds the workflows with your team and trains you to run them. My one-liner for it is “I train, not retain.”
| AI marketing consultant | Agency | Software vendor | |
|---|---|---|---|
| Who does the work | Your team, with the consultant building and training alongside | The agency, on your behalf | Your team, inside the vendor’s platform |
| What you own at the end | The workflows, their written sources and review points, and a trained owner. I train, not retain. | The work the agency delivered | Access to the platform while you subscribe |
| Where people stay in the decisions | At review points agreed before the build: brief, draft and final | Wherever the agency’s process asks for your sign-off | Wherever the platform’s built-in approval flow allows |
| The right call when | You want your own team running AI-assisted work | You want the work done for you, not run in-house | One platform covers the work and its approval flow fits your team |
24 of 25
Google showed an AI Overview on 24 of the 25 brand strategy, brand consistency and AI-readiness searches we track, so the answer a buyer reads first is usually written by a model.
What does an AI consultancy do?
For a marketing team, an AI consultancy finds where AI earns its place in the work, grounds it in the team’s own context, builds the workflow and trains the people who will run it.
In my client delivery that has meant campaign brief preparation, recurring performance reporting and first drafts prepared from approved source material, each held for a person to review.
I offer it three ways: implementation, private team workshops and individual mentorship.
How do you choose an AI marketing consultant?
Choose on how they work, not on the tools they name. Five questions sort it quickly.
- What happens in the first two weeks? Look for an audit of your tools and workflows and a prioritized plan, not a tool demo.
- What will we own at the end? Look for workflows, written sources and a trained owner on your team, not an open-ended retainer.
- Where do people approve the work? Review should happen at set checkpoints (brief, draft and final), not only at the end.
- Does training happen on our real work? It should, rather than on generic prompts.
- What results do you promise? Results depend on the workflow and the team, so be wary of anyone promising time saved or revenue before seeing yours.
Who writes this, and where to hire me
I’m Adam Sandler. More than two decades in marketing shape how I approach this work, including digital strategy and eCommerce experience at Nestlé Health Science, Edelman, and American Express.
If you want this for your own team, start with the workflow where work gets stuck. Tell me about it and we will agree the right starting point.