- name:
- creator-led-growth-90-days
- description:
- Use this skill when creator marketing should become a program, not an experiment — "make creators a real channel," "scale from a test to an engine," "what does quarter one of creator-led growth look like." Three 30-day phases: build the first cohort with tracking, launch–measure–cut, then systematize into a documented, forecastable engine with a second cohort matched to the winner profile.
Use when building the creator channel as a quarter-long program. Produces, by day 90: a documented playbook, a per-creator leaderboard, and a pipeline forecast built on real cost-per-qualified-click data.
Days 1–30 — foundation and first cohort
- week 1: define the ICP and core message with precision
- week 2: source 8–12 creators at 70–90% audience-ICP fit; follower count is secondary
- week 3: set commercial terms and wire tracking (qualified click = tracked + ≥30s on-site)
- week 4: brief with angles and proof points — never scripts
- milestone: briefed cohort, live tracking
Days 31–60 — launch, measure, cut
- stagger posts over two weeks so early angle signal shapes the rest
- read qualified-click data per creator; ignore vanity metrics — reach without relevance produces clicks that don't convert
- by week 8: cut the bottom third on data, not sentiment, and redeploy budget to the top performers
- milestone: per-creator leaderboard, budget reallocated
Days 61–90 — systematize into an engine
- document what won: creator profiles, angles, post structures
- build the pipeline forecast from your own cost-per-qualified-click history
- launch the second cohort matched to the proven winner profile
- milestone: playbook, defensible forecast, repeatable process
Expect concentration throughout — the top third of creators generates the majority of qualified clicks. The program's job is finding them fast and doubling down.
What good looks like
A great quarter ends with a document another operator could run: named winner profiles, tested angles, real unit costs, and a forecast finance accepts. A mediocre one ends with "we tried some creators" and no reusable knowledge. The overlooked failures: skipping the cut (sentiment keeps underperformers funded), scripting posts (destroys the peer trust being paid for), and reporting impressions to leadership when qualified clicks are the only number that compounds.
MUST cut the bottom third by week 8 on data. MUST document winner profiles and angles before scaling. NEVER report vanity metrics as program results.