How to Upskill Your Team for AI: A 90-Day Plan

Upskilling a team for AI works best in stages over about 90 days: expose them to a real use case, build daily habits, then deepen skills with shared prompts and peer learning. Skills stick when practice is tied to actual work, not abstract courses. Trying to teach AI in the abstract, with general overviews and theory, rarely changes how people work, because there is nothing concrete for the learning to attach to. A staged approach fixes that by anchoring each phase in tasks people actually do: comfort first, then habit, then depth. Upskilling is also an investment in your people, not only your operations, since the ability to work effectively with AI is increasingly valuable across roles. This guide lays out a three-stage, 90-day plan with clear milestones, shows how to measure progress with simple signals, and explains why letting peers carry the learning spreads skill faster than any single class.

Why AI skills are worth building now

The WEF Future of Jobs report ranks AI and analytical thinking among the fastest-growing skills, so upskilling is an investment in your people as much as your operations.

Days 1-30: exposure

Run a hands-on session on one real task, set up approved tools, and get everyone to try AI on their own work. The aim is comfort, not mastery. Our guide to training a skeptical team covers this first stage.

Days 31-60: habits

Encourage daily use on routine tasks, start a shared prompt library, and hold short weekly check-ins to share what works. Habits form through repetition on real work.

Days 61-90: depth

Now build skill: better prompting, combining tools, and tackling more complex tasks with proper review. Let your AI champion lead peer learning so knowledge spreads.

Measure progress simply

• Share of the team using AI weekly.
• Size and use of the shared prompt library.
• Tasks now done faster or with fewer errors.
• Confidence, gauged by a quick pulse check.

Keep it grounded in real work

Abstract training fades. The NBER study of support work found gains came from applying AI to actual tasks, with newer staff benefiting most. Treat it as context, and keep every exercise tied to real jobs.

Upskilling spreads fastest through peers, not top-down instruction. When a colleague shares a prompt that saved them an hour, others adopt it because they trust the source and see the benefit directly. Build this in by giving the team easy ways to share: a living prompt library, short weekly check-ins where people show one thing that worked, and a champion who collects and tidies the best examples. Peer learning also builds resilience, because the knowledge lives across the team rather than in one person's head.

Recognize the people who help others. A quick public thank-you for sharing a useful prompt signals that teaching is valued, which encourages more of it. Over a few weeks this turns into a culture where improving how you work with AI is just part of the job, rather than a project that ends. Our guide on training your team covers building that culture, including bringing along the people who start out reluctant.

Keep skills fresh as tools change

AI tools update often, so upskilling is never quite finished. The aim is steady currency, not constant retraining: a light monthly share, a maintained library, and a champion who flags the few updates genuinely worth adopting. Judge new features by whether they improve a task you actually do, not by how impressive they sound. Broad workforce research such as the WEF Future of Jobs report emphasizes continuous learning as a growing need, which a light ongoing routine satisfies without overwhelming the team.