1 min lesson
The start-narrow playbook
Start at the first move in "The start-narrow playbook" and carry it through to the proof.
Step 1 of 2
The start-narrow playbookthe field team's own rollout pattern
Cursor's own guidance is blunt: don't roll out to everyone at once. The discipline is proving ROIReturn on Investment. The value gained versus what it cost, the language an economic buyer funds deals in. Press Enter for the full definition. on a narrow surface before widening.
- 1Start narrow. One repo, a few well-chosen tasks, a few engineers - not a broad launch.
- 2Prove on specific tasks. Show measurable impact against the day-zero baseline on those exact tasks, so the win is attributable.
- 3Expand to a couple of teams. Share the results upward and recruit the next champion before pitching the next team.
- 4Scale org-wide only on strong ROIReturn on Investment. The value gained versus what it cost, the language an economic buyer funds deals in. Press Enter for the full definition.. Company-wide rollout is the reward for proven impact, not the opening move.
When a team asks "what should we automate first?", the field-tested answer is a daily "summarize changes" automation that posts what shipped - a low-risk, high-visibility starter that drove strong retention (~89%) where teams adopted it.
It's the ideal phase-one artifact: it's safe (read-and-summarize, not edit-and-merge), it makes the pilot's progress visible to leadership for free, and it teaches the team the automation muscle before they reach for riskier ones.
Learn more
Full explanation
Diagnose adoption from the admin analytics, by quadrant
Diagnose adoption from the admin analytics, by quadrantturn the dashboard into next actions
Once the pilot is live, two analytics signals read together tell you exactly what to do next: request volume and acceptance rate of committed code. Don't read them in isolation - the combination names the intervention.
- Request volume
- High
- Acceptance rate
- High
- What it means → your move
- Power users → identify them, study how they prompt via shared knowledge/transcripts, and distribute those learnings
- Request volume
- High
- Acceptance rate
- Low
- What it means → your move
- Enablement gap → they're trying hard but output isn't landing; this is a training opportunity, not a churn signal
- Request volume
- Low
- Acceptance rate
- -
- What it means → your move
- Awareness gap → drive more adoption; the tool isn't yet in the daily loop
| Request volume | Acceptance rate | What it means → your move |
|---|---|---|
| High | High | Power users → identify them, study how they prompt via shared knowledge/transcripts, and distribute those learnings |
| High | Low | Enablement gap → they're trying hard but output isn't landing; this is a training opportunity, not a churn signal |
| Low | - | Awareness gap → drive more adoption; the tool isn't yet in the daily loop |
Admins can enable shared knowledge / shared transcripts at team level so you can study exactly how a power user works.