When I talk about using AI in marketing, I start with the work we need to accomplish. What question are we trying to answer? What decision needs to be made? Where is the team spending time that could be used more effectively?
Those questions give the technology a purpose. They also keep responsibility where it belongs: with the people doing the work.
Start with a marketing problem.
In my agency work supporting five luxury senior living communities, I used AI to support research, strategic planning and audience-specific targeting. It helped me explore considerations and refine possible approaches. I remained responsible for the decisions, brand direction and quality of the work.
That is the relationship I bring to AI advisory: use it to assist the thinking and execution, while an experienced person evaluates the output and its relevance to the business.
A useful first application might be organizing existing research, exploring content directions or preparing questions for a planning session. The right starting point depends on the team and the information available.
A plausible persona is not a researched persona.
AI can help structure a buyer persona or suggest questions to investigate. But a polished profile is not evidence that a real buyer thinks or behaves that way.
I want teams to separate what they know from what they assume. Customer interviews, prospect conversations, sales feedback and existing research give persona work its grounding. AI can help organize those inputs; it should not be used to disguise missing evidence.
If a profile includes a motivation or concern that has not been validated, treat it as a hypothesis to explore. That keeps a useful starting point from becoming an unsupported marketing decision.
Give the output a human review.
For AI-assisted content, I ask practical questions. Are the facts supported? Does the message sound like the organization? Does it answer something the intended audience cares about? Could a reader misunderstand what is being offered?
That review matters especially when several teams or specialists contribute to a campaign. Speed is useful only if the work remains accurate, coherent and relevant.
Clear ownership helps: someone should know who checks the facts, who approves the language and who decides whether the work is ready to use.
Build a workflow your team can repeat.
AI adoption becomes more useful when it moves beyond isolated experiments. Choose a defined task, agree on what good output looks like, establish what information is appropriate to use and identify the review required before publication or delivery.
Then evaluate whether the process is helping. Is it saving time? Improving clarity? Helping the team ask better questions? Creating more editing than it removes?
AI can support marketing. Human judgment gives the work its direction.
My goal is to help teams use AI with purpose, while preserving their understanding of the customer and their responsibility for the brand. The technology should make the work more useful—not make the decisions harder to explain.