synthesize-insights — Claude Code command: what it does
synthesize-insights turns scattered customer feedback and usage numbers into one brief for leadership: themes, quotes, metrics and owners.
Apache-2.0 ★ 450 updated 2026-04-03
.claude/commands/synthesize-insights.mdWhat synthesize-insights does
You point the command at a topic, a set of sources and the people who will read the result. --theme, --sources and --audience carry those, --format picks a deck, memo or digest, and --urgency sets how wide the pass goes.
Then it runs five stages. Signals get gathered and de-duplicated, observations are clustered by theme, sentiment and impact, and each theme picks up numbers: how often it appears, the ARR attached, the shift in retention. Recommendations with owners and next-experiment ideas close the loop, and the run ends with a brief, tagged repository entries and a follow-up tracker.
The synthesis itself lands with the subagents customer-insights-partner, segmentation-architect and retention-analyst, while the insight-repository and activation-map skills keep tags and history. This is an orchestration layer, so a project missing those pieces leaves the middle stages with nothing to run.
What it helps with
- pull interviews, surveys and community threads into a single brief
- quantify how much ARR a recurring complaint touches
- tag observations by theme, sentiment and impact level
- hand product and GTM owners actions plus next-experiment ideas
- produce a memo instead of a deck for an exec audience
Example prompts for the agent
- Synthesize onboarding feedback for product and marketing: `/customer-analytics:synthesize-insights --theme "onboarding" --sources "interviews,surveys,community" --audience "product,marketing"`
- Give me a rush digest on churn for the exec team:
/customer-analytics:synthesize-insights --theme "churn" --sources "NPS,usage" --audience "exec" --urgency rush --format digest - Turn the adoption survey into a memo for customer success:
/customer-analytics:synthesize-insights --theme "adoption" --sources "surveys" --audience "cs" --format memo
How to install synthesize-insights
Manually: put the files in place
.claude/commands/synthesize-insights.mdCall it with /synthesize-insights. For all projects use ~/.claude/commands/
Manually: put the files in place
~/.codex/prompts/synthesize-insights.mdCall it with /prompts:synthesize-insights. Codex marks this format deprecated in favour of skills
Manually: put the files in place
.github/prompts/synthesize-insights.prompt.mdManually: put the files in place
.gemini/commands/synthesize-insights.tomlGemini stores commands as TOML: the prompt text goes into the prompt field
Questions about synthesize-insights
How do I install synthesize-insights in Claude Code?
Save the command file to .claude/commands/synthesize-insights.md in your project. It then shows up as /customer-analytics:synthesize-insights.
What does synthesize-insights do?
It gathers the sources you name, groups observations by theme and impact, attaches metrics to each theme and adds recommendations with owners. You get an insight brief, tagged repository updates and a follow-up tracker.
Which flags does the command take?
--theme is the focus topic such as onboarding, adoption, support, expansion or churn, and --sources lists inputs like interviews, surveys, community, NPS or usage. --audience names the readers, --urgency takes normal or rush, and --format takes deck, memo or digest.
Source text
synthesize-insights.md · Apache-2.0
--- name: synthesize-insights description: Consolidates qualitative + quantitative customer signals into executive-ready briefs. usage: /customer-analytics:synthesize-insights --theme "onboarding" --sources "interviews,surveys,community" --audience "product,marketing" --- # Command: synthesize-insights ## Inputs - **theme** – focus topic (onboarding, adoption, support, expansion, churn). - **sources** – comma-separated inputs (interviews, surveys, community, NPS, usage). - **audience** – target stakeholders (product, marketing, sales, cs, exec). - **urgency** – normal | rush to tailor scope. - **format** – deck | memo | digest. ## Workflow 1. **Signal Collection** – gather requested data + qualitative notes; dedupe/resample as needed. 2. **Coding & Tagging** – cluster observations into themes, sentiments, and impact level. 3. **Quantification** – add metrics per theme (occurrence, ARR impacted, retention delta). 4. **Recommendation Layer** – map insights to actions, owners, and next experiment ideas. 5. **Packaging & Distribution** – produce deck/memo/digest plus repository updates. ## Outputs - Insight brief with themes, quotes, metrics, and action items. - Repository updates with tags + links. - Follow-up tracker for product/GTM owners. ## Agent/Skill Invocations - `customer-insights-partner` – leads research synthesis. - `segmentation-architect` – contextualizes insights by segment. - `retention-analyst` – quantifies impact on retention metrics. - `insight-repository` skill – maintains tagging + historical logs. - `activation-map` skill – links insights to GTM actions. ---
© gtmagents · Apache-2.0 · gtmagents/gtm-agents