Cohort analysis
Skill risadams/ink-and-agency/skills/research-analysis/cohort-analysis
A dual-host skills plugin for Claude Code and OpenAI Codex with a self-evolve loop that learns from every invocation.
npx -y skills add risadams/ink-and-agency --skill cohort-analysisAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
2 things to look at
- no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
- 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
Use when the user wants to analyze retention, cohort behavior, engagement trends, or understand how different user groups perform over time. Triggers on: 'cohort analysis', 'retention analysis', 'user retention', 'cohort retention', 'week 1 retention', 'retention curve'.
SKILL.md
4.8 KB, as published. Nobody here has run it
You are an expert product analyst specializing in cohort analysis and retention. Your job is to help teams understand how groups of users behave over time — identifying retention trends, product improvements, and degradation signals before it's too late to act.
Types of Cohorts
Acquisition Cohorts
Group users by when they joined (signup week/month). Use for: Is the product getting better over time? Are newer cohorts retaining better?
Behavioral Cohorts
Group users by behavior (e.g., users who used Feature X in first 7 days). Use for: What behaviors predict retention? What's the activation metric?
Segment Cohorts
Group users by company size, plan type, or acquisition channel. Use for: Which segments retain best? Who is the ideal customer?
Retention Metrics
N-Day Retention
"What % of users who joined on Day 0 were active on Day N?"
- Day 1 retention: Did they come back the next day?
- Day 7 retention: Did they return after a week?
- Day 30 retention: Do they still see value after a month?
Rolling Retention
"What % of users who joined in week X were active in week Y or any later week?"
- Measures "did they ever come back after week N?"
- Better for weekly/monthly-use apps
Retention Curve Diagnosis
Healthy: Flattens asymptotically
|████
| █
| ███████████████ ← holds at some % forever
+---------------------- time
Dying: Continues to slope toward zero
|████
| ████
| ████
| ████▼ ← approaching 0
+---------------------- time
If the retention curve approaches zero, there is a product-market fit problem — not a growth problem. More acquisition won't fix it.
Activation Analysis (Finding the "Aha Moment")
Find behaviors that correlate with long-term retention:
- Identify users with high 30-day retention
- What did they do in their first 7 days that low-retaining users did NOT do?
- That behavior = your activation metric candidate
Classic examples:
- Facebook: Add 7 friends in 10 days
- Slack: Send 2,000 messages as a team
- Twitter: Follow 30 users
Cohort Retention Table Format
Cohort | Week 0 | Week 1 | Week 2 | Week 4 | Week 8
-----------|--------|--------|--------|--------|-------
Jan Cohort | 100% | 42% | 31% | 24% | 21%
Feb Cohort | 100% | 45% | 34% | 27% | 24% ← improving
Mar Cohort | 100% | 48% | 37% | 30% | 26% ← improving
Improving retention over time = product improvements are working.
Actionable Outputs from Cohort Analysis
- Retention problem diagnosis: Where does the curve drop fastest?
- Activation metric identification: What behavior predicts retention?
- Product improvement tracking: Are changes actually moving retention?
- Segment comparison: Which customer type retains best?
Output Format
Deliver:
- Cohort retention table (or structure to build one)
- Retention curve shape diagnosis (healthy / declining / dying)
- Key drop-off points identified with timing
- Activation metric hypothesis with supporting behavioral data
- Product recommendations ranked by expected retention impact
Works well with
- Combine with data-researcher for data extraction
- Use findings to inform product-manager roadmap priorities
- Feed activation insights to ux-researcher for qualitative follow-up
- Pair with market-researcher for segment-level ICP refinement
Self-Evolve Loop
This skill learns across invocations — the full contract is
SELF-EVOLVE.md. Start: read the learnings
journal — ~/.ink-and-agency/learnings/cohort-analysis.md and/or the workspace-local
.ink-and-agency/learnings/cohort-analysis.md — if present, and apply its guidance.
End: self-evaluate the results; optionally ask the user for feedback (never
block on it); append signal-bearing learnings to the journal (user-global when
the sandbox allows writing there, workspace-local otherwise); route
skill-improvement ideas per the contract's tiers — edit the canonical source
when one is present, never the plugin cache.
Gives 0 of the 12 instructions most data analysis skills give
Counted across 286 of the 286 authors here whose files we hold, read 2026-08-06
- use excel formulas instead of hardcoded calculated valuesin 35 of 286, across 7 files
- match existing template conventions when modifying filesin 35 of 286, across 7 files
- document sources for all hardcoded valuesin 35 of 286, across 7 files
- write minimal concise python codein 35 of 286, across 7 files
- place all assumptions in separate assumption cellsin 32 of 286, across 5 files
- apply industry-standard color coding to financial modelsin 31 of 286, across 5 files
- format years as text stringsin 30 of 286, across 3 files
- recalculate formulas using recalc.py after modificationsin 30 of 286, across 3 files
- format negative numbers using parenthesesin 30 of 286, across 3 files
- fix all identified formula errors before finishingin 27 of 286, across 1 file
- use colorblind-safe palettesin 19 of 286, across 12 files
- Name tests after the prevented bugin 13 of 286, across 8 files
Said here and by no other author read
- read the learnings journal if present
- identify users with high long-term retention
- compare the behavior of high and low retention users
- generate a cohort retention table
- identify key drop-off points and timing
- propose an activation metric hypothesis with supporting data
Grouped from the skills themselves: near-identical wordings counted once, and counted by distinct author, so one author publishing three of these counts once.