Menu
You put the room together because the company needs a shared baseline.
Then the room tells the truth.
Someone asks where the AI hub lives. Someone else wants a walkthrough of migrating work they already built. A third person is quietly bored because they passed "where is the button" months ago. Leadership wanted one session. The skill gap wanted three.
That isn't mostly a curriculum problem. It's a mixed-skill room: one agenda asked to serve people who are still finding the door and people who are already building inside.
If you train as if everyone is in the middle, beginners leave lost and power users leave untouched.
Mixed skill is normal in mid-market rollouts. It becomes a problem when design pretends it isn't.
Beginners need orientation: where approved tools live, what is safe to paste, what a first useful job looks like on their desk. Intermediate people need demonstration on work they already own. Power users need standards, critique, and a path to share what they built so it doesn't stay private forever.
One generic workshop usually optimizes for none of those groups. It talks features at a pace that feels slow to builders and fast to people who still can't find the hub. Feedback forms can still look warm. Capability does not even out.
Pace anxiety shows up here too. Teams that feel "so slow" are often comparing their beginners to someone else's power users. The fix is honest level design, not louder urgency.
We keep seeing the same picture without needing a named account to make the point.
A leader names the challenge out loud: people are at different skill levels, and even the sponsor still needs a clearer grasp of the basics before advanced moves. In the same arc, someone asks for a walkthrough and where to find the hub. Elsewhere, a regional leader discovers a capability midstream and says they had no idea you could do that. Another operator feels behind on pace while the room is still uneven underneath.
That is one problem, not a talent shortage. You don't need a harder workshop first. You need to know who's in which band before you design the room.
Sometimes for a short shared baseline. Rarely for the whole arc.
A short shared open can give everyone the same vocabulary: approved tools, data rails, what "good" means on company work. After that, split.
Forcing mixed skill through a single advanced demo teaches the wrong people twice: beginners memorize steps they can't yet place, and builders never get the governance and handoff work they actually need.
Don't guess from titles. Ask about behavior on real work.
Useful pre-checks:
Answers sort people faster than a self-rating of "beginner / intermediate / advanced." Self-ratings flatten. Behavior on owned work doesn't.
If you skip the pre-check, you will rediscover the mix in the first twenty minutes of the session, when it is already too late to redesign the room.
Reinforcement is not the tip deck again.
After a first session, people return with friction from real attempts: permissions, unclear inputs, uncertainty about what is safe, a path that worked once and failed on the second try. Reinforcement should start from those attempts.
A useful second session covers:
If the second session introduces a brand-new feature tour, you restarted novelty. You did not reinforce capability.
Should beginners and power users be trained together?
For a short shared baseline on rails and vocabulary, yes. For the full enablement arc, usually no. Beginners need orientation and one supported job. Builders need critique, standards, and handoff discipline. A single advanced demo through a mixed room leaves both groups underserved.
What should a reinforcement session cover that the first one didn't?
Friction from real attempts on the same path: what failed between sessions, decisions people still can't make alone, and how builders share without shadow systems. Reinforcement starts from application, not from a new feature tour.
How do you find out who's at which level before training?
Ask about behavior on owned work: can they find the tools, have they used AI on a recurring job, can they explain keep vs throw away, have they handed work off, are they already building private workflows. Behavior sorts levels faster than self-ratings or job titles.
If the pattern on this page sounds familiar, start with the free AI Operations Reality Check. One everyday workflow. No pitch deck.
If you want a bounded diagnostic and a written next-step memo, the AI Operations Review is the paid step after that ($997, not an open-ended contract).
Susan Westwater is the CEO and Co-Founder of Pragmatic Digital. She helps mid-market and PE-backed teams move from scattered AI pilots to governed, measurable workflows that actually deliver operating leverage. With 25+ years in CX and brand leadership at Leo Burnett and Ricoh USA, Susan specializes in turning AI ambition into repeatable systems that protect brand voice and reduce revision cycles. She is co-author of Voice Strategy and Voice Marketing.
Weekly insights on AI adoption, workflow, and what's actually working in mid-market organizations.