Running the Program
You have the philosophy, the tiers, the rubric, and a way to measure adoption. None of it moves until someone runs it. This page is the operating manual: how to put the framework into motion and keep it there.
The work falls into four moves, in order: find your champions, choose your assessors, run the conversations, and set a cadence that survives a busy quarter. Treat it as a loop, not a launch. You'll cycle through it as people climb.
The operational loop
Identify champions
Find the few people who already use AI well and want others to. They seed the program and model what "good" looks like.
Choose assessors
Pick people who can judge reasoning, not recitation. They run the conversations and keep scoring honest and consistent.
Run the conversations
Assess through dialogue against the rubric. Score Solid / Partial / Gap, and say "Not Yet" where the depth isn't there.
Set a cadence
Decide how often people advance and how often you re-measure. A predictable rhythm beats a heroic one-time push.
1. Identify champions
Every program needs people who are already bought in: not appointed, but genuinely excited. These are your champions. They're usually the ones quietly building useful things before anyone asked them to.
Champions do three jobs. They model what capable looks like, so the rest of the team has a reference point. They answer the small questions that would otherwise stall a beginner for an afternoon. And they give you honest signal, telling you where the framework is working and where it's friction.
Pick champions who are respected by their peers, not just senior. Influence travels through trust, and a credible mid-level engineer often moves a team faster than a manager mandate.
Keep the group small to start. Two or three is plenty for a first cycle. You want depth and enthusiasm, not a committee. Give them explicit permission to spend time on it; "be a champion" without protected time is just extra unpaid work.
2. Choose assessors
Assessors are the people who run the assessment conversations and decide where someone lands. The choice matters more than any other operational decision, because the assessor is the rubric in the room, and a vague assessor produces vague results.
Good assessors share a few traits:
- They can hear reasoning. They listen for why someone made a choice, not whether they recited the right keyword.
- They're calibrated against each other. Two assessors should land in roughly the same place on the same person. Without calibration, a tier means something different depending on who you drew.
- They're comfortable with "Not Yet." Saying someone isn't there yet is a kindness, not a failure, but only if the assessor can deliver it without flinching.
Don't let someone assess their own direct reports if you can avoid it. The reporting relationship distorts both the questions and the answers, and it puts the rubric's credibility at risk.
Calibrate before you go live. Have your assessors score the same example together and talk through where they disagree. That single exercise does more for consistency than any written guideline.
You don't need a big roster to start. Begin with a small core, two or three people you trust to hold the bar, because a tight group is far easier to calibrate. The pool then grows itself: as people come through the conversations, the strongest of them become your next assessors. Each cycle widens the bench, and a wider bench means more conversations running in parallel: more throughput, more speed, and adoption that spreads on its own momentum instead of bottlenecking on the handful of people who started it.
When someone passes a tier with flying colors, full Solids across the board, don't just congratulate them. Ask them to assess the next person, or make it a condition of the result. Every strong pass becomes another qualified assessor, and the program cascades through the organization rather than stalling on its original few. That's where the real leverage comes from.
3. Run the conversations
This is where the framework meets a real person. The assessment is a conversation, not a quiz — you're probing for judgment and depth, then mapping what you hear onto the rubric's Solid / Partial / Gap scale.
Three things keep the conversations honest:
- Score against the rubric, every time. The rubric exists so that "Practitioner" means the same thing in March and in September. Drift starts the moment people score from memory.
- Use "Not Yet," not "Fail." A gap is a direction, not a verdict. Tell the person what they'd need to demonstrate to clear the bar; that turns a hard result into a plan.
- Capture evidence, not just a number. A one-line note on why a dimension scored the way it did is what lets the person — and the next assessor — act on the result.
The detailed mechanics of facilitating a single conversation — the questions, the follow-ups, the room — are their own craft. This page is about running the program; the per-conversation technique is something your assessors develop with reps and calibration.
4. Set a cadence
A program is only as good as its rhythm. Decide up front how often people can advance and how often you re-measure adoption, then protect those dates like any other commitment.
A workable starting cadence:
- Assessment windows every quarter, so people always know when the next opportunity is and can prepare.
- An adoption re-measure on the same quarterly beat: revisit The Climb and check whether usage is deepening, not just whether seats went up.
- A short champions sync monthly, so friction surfaces early instead of at the next big review.
Start slower than feels ambitious. A cadence you can actually hold for a year builds more capability than an aggressive schedule that quietly lapses after two cycles.
Then close the loop. The results from one cycle tell you who your next champions are, where the rubric needs sharpening, and where adoption stalled. Feed that straight back into the next round. The program compounds when each cycle makes the next one better.
Take it and run
That's the whole machine: champions to seed it, assessors to judge it, conversations to measure it, and a cadence to keep it honest. It's a real framework, and it's yours — take it and run with it.
And if you'd rather not do it alone — standing this up is real work, and a second set of hands who've run it before can save you a few false starts — Jahnel Group can help. No pressure either way. The framework stands on its own.