// resposta · ia no marketing
How do I prepare my marketing team to use AI?
Atualizado em 22 de julho de 2026
Preparing a marketing team for AI takes three moves: redefining roles, training by doing together on real routine tasks, and deciding upfront what will never be delegated, like strategy, brand voice and final approval. AI takes on raw execution; people take on direction and judgment.
The roles that change
AI doesn't eliminate jobs, it shifts where the effort goes in each one:
- Copywriter becomes editor and prompt director: writes the brief, guides the AI and reviews the output with judgment.
- Designer becomes an image curator: generates variations, picks and refines with real brand assets.
- Media analyst uses AI to read reports, build test hypotheses and gain speed in analysis.
- Customer service gets reply drafts, but the person actually talking to the client stays human.
What doesn't change: strategy, knowledge of the client and responsibility for what goes out under the company's name.
The adoption process that works
The training that works is doing it together, on a real task, in this order:
- Pick a real, repetitive task from the week: a post caption, an ad variation, a meeting summary.
- Run it together: someone from the team and someone who already knows the tool work side by side until the task comes out with quality.
- Document the process: the prompt that worked, the review pattern and what changed in execution time.
- Turn it into a routine: the task enters the workflow with a defined owner and quality standard.
- Only then move on to the next task.
A theoretical course with no real task turns into a forgotten note. When choosing tools, use the guideline in which AI tools are worth it for a small company. And the underlying question, whether AI will replace agencies, also applies inside your own company: AI eats up execution and raises the value of whoever directs it.
What not to delegate to AI
- Strategy decisions and where the budget goes.
- The brand's voice and positioning.
- Client relationships, especially negotiation and conflict.
- Final approval of everything that goes live.
- Sensitive data: never paste client information into a public tool without a clear policy.
In practice, in the countryside
In a mid-size company in the Northwest of São Paulo state, the marketing team usually has one to three people wearing multiple hats. In that setting, well-adopted AI takes weight off execution and gives back time for client visits, trade shows and fieldwork. At desigual, AI production runs on our own infrastructure, with an RTX 4090 and ComfyUI, and the person operating it is still someone with real judgment. The same principle applies inside your own company.
Want to structure AI adoption on your team with a team that already runs this in real production? Reach out on desigual's WhatsApp and let's set up a conversation.
Perguntas relacionadas
how do you train a marketing team to use artificial intelligence
The training that works is doing it together: pick a real task from the routine, run it with AI alongside the team, and document the process that worked. A theoretical course with no real task rarely changes day to day work.
what should AI not do in my company's marketing
Don't delegate strategy decisions, the brand's voice, client relationships, and final approval of what goes live. AI executes and speeds things up; direction, judgment and responsibility stay human.