Fix 'they bought it but nobody uses it.'
Reads what's unused, picks the one habit to build next, and writes the enablement plan that gets there.
Inside: The adoption gap, the next habit, and the enablement plan.
Install it in one line, or paste it in.
~/.claude/skills/ and runs automatically when it is relevant.Connect your context. Set it to your motion.
it reads real feature and seat usage automatically, across the whole account.
pulls the entitlement, the plan tier, and the account owner so the plan lands on the right desk.
shows what was already promised in onboarding, so the plan does not repeat a training they ignored.
This was built for a B2B SaaS product with multiple features and per-seat licensing. Set these to your stack:
Run any adoption model you like. The skill measures "what did they buy versus what do they touch," so point it at your core features, not anyone else's.
| Set this | What it is | Default / Example |
|---|---|---|
| ANALYTICS | your product-analytics connector | a product-usage tool of your choice |
| CRM | your CRM connector | your CRM of choice |
| CORE_FEATURES | the features that define value | your top 5 to 10 capabilities |
| ACTIVATION_EVENT | the action that means "adopted" | first workflow completedfirst invite sent |
| SEAT field | where entitlement lives | licensed seats vs active seats |
| IDLE_DAYS | no-login days that mean a seat is idle | 30 (re-tune to your cycle) |
| ADOPTION_TIER | how you rank a user's depth | Power / Established / Emerging / Dormant |
Everything the skill does, in full.
Reads how an account actually uses your product, finds the gap between what they bought and what they touch, and picks the single next habit worth building. Then it writes the enablement plan to get there: who to reach, what to show them, and how you will know it worked. One habit at a time, because adoption is a sequence, not a launch.
- 1Adoption gap (bought vs touched)
List what the account is entitled to. List what they actually use. The gap between the two is the whole job. Name the specific unused feature or the count of idle seats. Never say "low adoption" without the line item behind it.
- 2Pick the one next habit
Do not hand an account a ten-item training plan. Rank the unused features by value-to-effort: high value, low friction, adjacent to something they already do. The winner is the next habit. One. The rest wait their turn.
- 3Depth tiers per user
Sort the named users into tiers by how deeply they use the product (for example Power, Established, Emerging, Dormant). A Dormant user needs a different push than a Power user missing one feature. The plan targets the tier, not the average.
- 4Enablement plan (who, what, proof)
For the chosen habit, write the play: the named user or role to reach, the specific thing to show them, the reason it matters to their job, and the usage signal that will prove it stuck. If a champion drives adoption, name them.
- 5Momentum check
Adoption is a trend, not a snapshot. Compare usage over the trailing weeks. A feature used once and dropped is not adopted. Flag the difference between a spike and a habit.
- No "low adoption" verdict without the named feature or idle-seat count that proves it.
- Exactly one next habit per plan. If everything is urgent, nothing gets adopted.
- A feature counts as adopted only on a repeated signal, never a single event.
- Idle-seat claims show the last-login evidence, never a guess.
ADOPTION PLAN, one account Entitled: 40 seats, 9 core features In use: 26 active seats, 4 features touched Gap: 14 idle seats, 5 features never opened Next habit: shared templates (high value, low friction, next to what they already do) Target: the ops lead + 3 Emerging users Show them: one 15-min walkthrough on their own workflow Proof: 3+ templates saved and reused within 2 weeks Then, not now: - Reporting module (needs the template habit first) - The 14 idle seats (re-provision after the ops team is live)
IDLE_DAYS (30) and the tier cutoffs are defaults, not laws. They suited a product with a monthly usage rhythm. If your product is touched daily, tighten them. If it is a quarterly tool, loosen them. The logic does not change. The thresholds are yours.
Where an operator takes this next.
The one-habit plan works by hand on one account. Here's the always-on version.
One habit at a time, run automatically across every account instead of just the one you happened to look at.
Connect Amplitude or your product-analytics tool to a scheduled Claude task and get the adoption gap recalculated for every account before the CSM team logs on.
Have the task watch for a feature used once and dropped, and post the flag to Slack before it shows up as churn risk three months later.
Feed the "next habit" target list into HubSpot or your CRM so the named ops lead gets the walkthrough invite without a CSM having to remember.
One skill is the on-ramp.
A single skill does one job. Chained into a playbook, or run as a full build, it becomes a system. Here is where this one plugs in.