Most marketing leaders will tell you their team is using AI. But there’s a difference between a team that uses AI and a team that’s getting better at it together and sharing their progress.
Often, there’s a handful of people on the team who have genuinely figured out how to work with AI effectively. They’ve built smarter prompts, developed better workflows, learned how to feed the tools the right context to get consistently useful output. They’re moving faster. But almost none of what they’ve learned has transferred to anyone else on the team.
The distance between the two groups can keep growing if there’s no system to close it.
Paul Roetzer put it plainly in episode 221 of The Artificial Intelligence Show: in a team of 100 people with AI access, there are typically five to ten power users making daily breakthroughs. And then there’s everyone else. The problem isn’t the technology. It’s that learning isn’t being treated as a team asset.
What To Do About It Now
A few places to start:
Find your power users and make their workflows visible. Identify the people whose AI output consistently impresses, who finish things faster, who seem to get more out of the tools. Ask them to document what they’re actually doing. Not a polished tutorial. Just a working description of how they’ve structured their prompts, what context they’re feeding in, what they’ve learned about what works.
Build shared prompt and project libraries. If one person has built a well-structured AI project for writing email campaigns that consistently produces on-brand output, that project should be accessible to the whole team, not living in their personal account. Shared resources like this are a practical way to transfer the benefit of one person’s learning to everyone.
Create a lightweight feedback loop. A simple, recurring practice — even fifteen minutes at a team meeting — where someone shares one AI workflow win or experiment that week changes the dynamic. It signals that this kind of learning is valued, surfaces things that would otherwise stay buried, and gives people specific things to try rather than vague encouragement to use AI more.
Treat context-building as a team project. The brand guidelines, audience personas, messaging frameworks, and campaign learnings that make AI output actually useful are team assets. Centralizing them — and making them easy to load into AI tools — multiplies the value of every person on the team who uses them.
The Compounding Problem
The people who are already ahead keep getting better faster. Why? Because they’re using AI more and learning more from each use. The people who aren’t plugged into that learning cycle stay roughly where they are.
Over time, the difference grows. Marketing leaders who get ahead of this now — by building even basic systems to share learning and capture context — are setting up their teams to build and advance together.
This blog is based on insights from Episode 221 of The Artificial Intelligence Show, hosted by Paul Roetzer and Mike Kaput.
Want to learn more about applying AI ochestration and AI agents to your workflows? Join us at our free virtual B2B Marketers Summit on June 25, 2026.
