Install any skill in seconds. Free to start, no credit card required.
Get Started Free →30-60-90 day plan for a new PM joining a company or team, grounded in Michael Watkins' First 90 Days framework and the STARS situational diagnosis. Includes week-by-week plan, stakeholder map, 1:1 question bank, and first-PRD template.
.claude/skills/borghei-pm-onboarding/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-05 | ✗→✓ | ▲ Improved | 68% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 64% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 69% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 75% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -13% | 0% |
A new PM's first 90 days are disproportionately important. Trust earned in the first quarter compounds; missteps in the first quarter haunt for the next year. This skill is a structured 30-60-90 day plan that helps a new PM diagnose the situation, build relationships, identify early wins, and arrive at the end of the quarter with credibility and a clear point of view.
The skill draws on Michael Watkins' The First 90 Days, the STARS situational diagnosis (Start-up / Turnaround / Accelerated growth / Realignment / Sustaining success), and the public PM onboarding patterns popularized by senior product leaders.
Diagnose the STARS situation first — a Sustaining-success area rewards 60+ days of listening; a Turnaround needs decisive action within 30. Misreading the situation is the most common new-PM mistake.
Before building the onboarding plan, confirm these inputs. If any is unknown or vague, ASK — do not assume:
Stop rule: ask only the 2-3 that most change the output. If the user says "just draft it," proceed and list your assumptions at the top of the artifact.
Load the reference that matches the task; keep this file lean and pull detail on demand.
assets/30_60_90_plan.md — editable 30-60-90 plan template.assets/stakeholder_map.md — onboarding stakeholder map template.assets/onboarding_1on1_questions.md — question bank for the first 30 days of 1:1s.assets/first_prd_template.md — tighter PRD template for new-PM constraints.External: Watkins, M. The First 90 Days (HBR Press, 2013); Bock, L. Work Rules! (onboarding at scale).
In Scope: 30-60-90 day plan for new PM roles; STARS situational diagnosis; onboarding stakeholder mapping; first-quarter 1:1 question banks; first-PRD template; early-win identification and execution.
Out of Scope: Job-search or offer-evaluation prep (use personal-productivity/); compensation negotiation; long-term career planning beyond 90 days (pm-career-ladder/); onboarding for non-PM roles.
Caveats: The 90-day plan is a planning artifact, not a contract — adjust as you learn. Calibrate "early wins" to the company's pace (a startup's 90-day win may be an enterprise's 180-day win). Your manager's expectations matter most: align in week 1, do not surprise them in week 6.
| Integration | Direction | What Flows | |-------------|-----------|------------| | pm-1on1s/ | Feeds into | The 1:1 question banks here are designed for first-90-day conversations; ongoing 1:1s use the broader skill | | pm-career-ladder/ | Feeds into | End-of-90-day retro becomes the baseline self-score on the ladder | | pm-interview-prep/ | Receives from | Pre-offer "Can you do the job?" stories often map to early wins delivered in past 90-day windows | | senior-pm/stakeholder-mapper/ | Reuses | The stakeholder mapping technique scales beyond onboarding into steady-state | | execution/create-prd/ | Feeds into | First-PRD template is a tighter version of the full PRD skill | | discovery/interview-synthesis/ | Reuses | Customer interviews in week 5 use the same synthesis discipline as steady-state discovery |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 17,268 | 32,730 | +90% | 1 | 1 | 0% | 2,684 | 6,463 | +141% | 0 | 0 | — |
case-02 | fail→fail | 23,140 | 24,187 | +5% | 1 | 1 | 0% | 3,318 | 4,909 | +48% | 0 | 0 | — |
case-03 | fail→fail | 16,944 | 19,185 | +13% | 1 | 1 | 0% | 2,463 | 4,381 | +78% | 0 | 0 | — |
case-04 | pass→pass | 16,242 | 9,142 | -44% | 1 | 1 | 0% | 2,446 | 2,866 | +17% | 0 | 0 | — |
case-05 | fail→pass | 13,733 | 12,172 | -11% | 1 | 1 | 0% | 2,011 | 3,369 | +68% | 0 | 0 | — |
case-06 | fail→pass | 15,191 | 13,841 | -9% | 1 | 1 | 0% | 2,230 | 3,655 | +64% | 0 | 0 | — |
case-07 | pass→pass | 16,513 | 15,100 | -9% | 1 | 1 | 0% | 2,499 | 3,782 | +51% | 0 | 0 | — |
case-08 | fail→pass | 12,844 | 12,472 | -3% | 1 | 1 | 0% | 1,966 | 3,323 | +69% | 0 | 0 | — |
case-09 | pass→pass | 14,056 | 12,244 | -13% | 1 | 1 | 0% | 2,167 | 3,431 | +58% | 0 | 0 | — |
case-10 | pass→pass | 16,480 | 17,588 | +7% | 1 | 1 | 0% | 2,594 | 4,171 | +61% | 0 | 0 | — |
case-11 | fail→pass | 13,019 | 12,823 | -2% | 1 | 1 | 0% | 1,929 | 3,384 | +75% | 0 | 0 | — |
case-12 | pass→pass | 12,579 | 10,914 | -13% | 1 | 1 | 0% | 1,927 | 3,154 | +64% | 0 | 0 | — |
case-13 | pass→pass | 12,947 | 7,713 | -40% | 1 | 1 | 0% | 1,880 | 2,617 | +39% | 0 | 0 | — |
case-14 | fail→pass | 19,971 | 5,772 | -71% | 1 | 1 | 0% | 2,745 | 2,383 | -13% | 0 | 0 | — |
case-15 | fail→pass | 20,218 | 7,039 | -65% | 1 | 1 | 0% | 3,099 | 2,636 | -15% | 0 | 0 | — |
case-16 | fail→fail | 21,280 | 26,328 | +24% | 1 | 1 | 0% | 3,123 | 5,530 | +77% | 0 | 0 | — |
case-17 | fail→pass | 10,235 | 7,377 | -28% | 1 | 1 | 0% | 1,625 | 2,701 | +66% | 0 | 0 | — |
case-18 | pass→pass | 9,178 | 5,675 | -38% | 1 | 1 | 0% | 1,400 | 2,388 | +71% | 0 | 0 | — |
case-19 | fail→pass | 14,872 | 11,035 | -26% | 1 | 1 | 0% | 2,090 | 3,205 | +53% | 0 | 0 | — |
case-20 | fail→pass | 14,628 | 12,059 | -18% | 1 | 1 | 0% | 2,344 | 3,293 | +40% | 0 | 0 | — |
case-21 | fail→fail | 10,853 | 8,670 | -20% | 1 | 1 | 0% | 1,530 | 2,906 | +90% | 0 | 0 | — |
case-22 | pass→pass | 16,320 | 11,301 | -31% | 1 | 1 | 0% | 2,282 | 3,298 | +45% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +41 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.