# Creator Programme Qbr > Use this skill when a quarter of a B2B LinkedIn creator programme needs reviewing for client leadership. Synthesises campaign performance into a strategic review — what happened, why, and what to do next — with Scale / Adjust / Stop recommendations. - Skill: `swan-gtm/creator-programme-qbr` (Agent Skill) - Install (CLI): `npx skillmds@latest add swan-gtm/creator-programme-qbr` - Raw SKILL.md: https://api.skillmd.com/api/skills/swan-gtm/creator-programme-qbr/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Marketing & Growth - Author: swan-gtm (https://skillmd.com/u/swan-gtm) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/swan-gtm/creator-programme-qbr --- # Draft Quarterly Programme Review Tell the client's leadership what their creator programme did, why, and what to change next quarter. This is a strategic document, not a stack of weekly updates: it reads top-down, with the story and the decisions first and the numbers underneath as support, and it's as clear about what to cut as what to grow. ## Connectors for this skill Optional. The skill runs from pasted inputs, but will use these if available: - **Your campaign performance records** (spreadsheet, CRM, or analytics store) — to pull the quarter's per-creator and programme-level metrics. - **A LinkedIn or creator-analytics source** — for audience demographics behind ICP reach. - **A document tool** — to produce the polished, client-ready deliverable. If none are connected, gather the inputs manually in Step 0. ## Step 0 – Gather the inputs Collect, asking the user for anything missing: - The client and the quarter under review. - The quarter's campaign performance: per-creator and programme-level impressions, engagement (likes / comments / reposts), link clicks, spend, and any conversion data (leads, cost-per-lead). - The prior quarter's equivalent figures, for comparison. - The pre-launch benchmarks or targets set for the quarter, if any. - The client's ideal-customer definition, so ICP reach can be reported. - Any client-side signals available (site traffic, brand search, inbound) for uplift. Guiding principles for the whole review: - **Decisions first, data second.** Lead each section with the call to make, then the numbers that justify it. - **No naked numbers.** Every metric sits against prior quarter, the benchmark, or the paid-media equivalent. - **Read the programme, not the posts.** Surface patterns across creators, niches, and formats. - **Every recommendation carries its receipt.** Tie each to a specific finding and the maths behind it. - **Be specific about what to cut.** A review that only ever says "do more" isn't a review. ## Step 1 – Pull and normalise the quarter's data Compute per-creator and programme-level figures. Use **medians** for rate and efficiency metrics (robust to spike days) and totals for volume (impressions, clicks, leads, spend). Pull the prior quarter the same way. Estimate ICP reach as `impressions × seniority-fit% × industry-fit% × company-size-fit%` using the client's ICP definition. ## Step 2 – Compare and find the patterns Attach a prior-quarter delta and a benchmark delta to every metric. Read the quarter through several lenses: creator performance, niche or theme, new vs returning creators, content format, and budget pacing. Identify the top and bottom creators and niches with the maths that proves it. ## Step 3 – Write the recommendations 4–7 recommendations, each tagged **Scale / Adjust / Stop**, each referencing a Step 2 finding and stating what it costs. Force a real Stop section. Then the planning inputs: next-quarter budget, roster targets (who to re-sign, add, or drop), KPI targets, key dates, and open decisions. ## Step 4 – Assemble the six sections 1. **Quarter snapshot** — a rollup table (this quarter vs prior vs benchmark) with a one-paragraph headline above it. 2. **Creator / campaign summary** — 3–5 lines each, ranked by performance. 3. **Trend analysis** — the lenses from Step 2. 4. **Wins and misses** — specific, each with its strategic implication; misses stated plainly. 5. **Strategic recommendations** — the Scale / Adjust / Stop set from Step 3. 6. **Planning inputs** — budget, roster, KPIs, dates, open decisions. ## Output Produce a polished, client-ready document (a branded doc, not a raw notes dump), proofed before hand-over. ``` [CLIENT] — Q[X] [YEAR] PROGRAMME REVIEW Headline: [one paragraph — the story of the quarter] 1. QUARTER SNAPSHOT This Q Prior Q Benchmark Impressions [ ] [ ] [ ] ICP reach [ ] [ ] [ ] Cost / 1k impressions [ ] [ ] [ ] Leads / CVR [ ] [ ] [ ] Spend vs budget [ ] [ ] [ ] 2. CREATOR SUMMARY (ranked) 3. TRENDS (by lens) 4. WINS & MISSES 5. RECOMMENDATIONS (Scale / Adjust / Stop) 6. PLANNING INPUTS (budget · roster · KPIs · dates · open decisions) ``` Use the user's own currency throughout. ## What good looks like - A client exec could read the headline alone and know the quarter's story. - Every metric carries a comparison; no naked numbers. - Every recommendation cites the finding behind it. - There is a real Stop or shrink call; if nothing underperformed, the review says why the whole roster earns its place, with the maths. - Synthesis, not summary: it doesn't read like a run of weekly updates stapled together. - No invented numbers: a metric that wasn't tracked is named as a gap with a recommendation to instrument it next quarter.