meta-cohort-analysis

name	meta-cohort-analysis
description	Use when comparing acquisition or retention cohorts using dated behavioural data. Produces cohort analysis with qualified findings; use `meta-reporting` when that neighbouring contract is the closer match.
metadata	{"portable":true,"compatible_with":["claude-code","codex"]}

Cohort Analysis for Client Reporting

Source: Raaz (c.2023) Web Analytics Blueprint

---

Use When

• Use this skill for comparing acquisition or retention cohorts using dated behavioural data.
• Confirm that meta-reporting is not the closer route before proceeding.

Do Not Use When

• Use meta-reporting when its narrower output is requested.
• Do not publish, spend, change a live account, certify compliance, or invent missing client evidence.

Required Inputs

Artefact	Source/provider	Required?	If absent
Event-level cohort data, date range and cohort definition	Client, approved systems, or dated platform exports	Yes	Stop the affected decision; request it or mark the field unknown and narrow the output.
Purpose, audience and approval boundary	Client brief or accountable owner	Yes	Return discovery questions; do not infer approval.

Outputs

Artefact	Consumer	Acceptance condition
Cohort analysis with qualified findings	Client lead and next workflow owner	Every recommendation traces to an input, names an owner or next action, and marks assumptions and unassessed checks.

Evidence Produced

Evidence	Format	Acceptance condition
Decision and source register	Table in the deliverable	Each material claim records its source/date or is labelled unverified; missing evidence never becomes a pass.

Capability and permission boundary

Read and search access to the supplied artefacts are required; calculation or file-rendering capability is optional. This is read-only by default: inspect and report without changing source records, accounts, skills or campaigns. Editing the deliverable requires explicit authorisation; publishing, production mutation, destructive action, spend, and certification claims require separate explicit authority and evidence.

Degraded mode

If files, platform access, network, rendering, fonts, or calculation tools are unavailable, return the narrowest useful qualified cohort analysis with qualified findings. Mark each blocked check not assessed, state the consequence, and provide the exact evidence needed to resume. Never convert an unavailable check into a pass.

Decision rules

Choice	Action	Failure or risk avoided
Event-level cohort data, date range and cohort definition is current and attributable	Produce the full cohort analysis with qualified findings and cite the evidence used.	Decisions based on stale or unrelated evidence.
A material input is missing or contradictory	Stop that decision, request clarification, or issue a labelled partial result.	Fabricated precision and false confidence.
The requested outcome belongs to meta-reporting	Route there and hand over the verified inputs already collected.	Neighbour collision and duplicated work.

Workflow

1. Confirm the requested decision, consumer, market, period and permission boundary; route to meta-reporting if its contract is closer.
2. Inventory the required inputs and their provenance. Stop any decision whose critical evidence is absent; recover by requesting it or recording a bounded assumption.
3. Apply the domain method in the core sections below, following the decision table whenever evidence conflicts or scope changes.
4. Verify calculations, dates, named platforms and claims against the supplied sources; label inference and uncertainty.
5. Produce the cohort analysis with qualified findings, decision/source register and explicit next owner. Do not mutate live systems without separate authority.
6. Run the repository anti-slop ship gate. If a blocking factual, permission or evidence defect remains, fix it or withhold release.

Quality Standards

The output is client-specific, uses British English and the stated market/currency, distinguishes observed fact from inference, exposes gaps, and gives a checkable acceptance condition. Recommendations must be feasible within the confirmed budget, capacity and permissions.

Anti-Patterns

• Using an undated benchmark as the client's result. Fix: use account evidence or label the benchmark as a provisional comparator.
• Producing the cohort analysis with qualified findings without event-level cohort data. Fix: stop the affected decision or issue a clearly bounded partial output.
• Treating missing access or data as a successful check. Fix: record not assessed, its risk and the recovery input.
• Absorbing meta-reporting into this workflow. Fix: route the neighbouring output and hand over verified inputs.
• Publishing, spending or editing a live account during planning or review. Fix: obtain separate explicit authority and retain action evidence.

Worked example

Given verified event-level cohort data, the skill produces a cohort analysis with qualified findings with source dates and named assumptions. If that evidence cannot be accessed, it returns only the supported sections plus a recovery list; it does not fill gaps with East African defaults.

Read next

• meta-reporting for the neighbouring contract.
• anti-ai-slop during production.
• ai-slop-audit at the release checkpoint.

References

• Anti-AI slop production gate
• Follow the directly linked repository skills above and any domain references named in the core sections below. Verify current platform, price, legal and regulatory claims before use.

Required Inputs

Ask for the following before generating any deliverable:

1. Client business name
2. Industry (e-commerce, services, B2B SaaS, hospitality, etc.)
3. Country / city (defaults to Uganda / East Africa)
4. Primary goal (e.g. demonstrate campaign ROI, identify best acquisition channel, justify budget reallocation)
5. GA4 access level (Admin / Editor / Viewer — determines which steps are available)
6. Reporting period (weekly or monthly cohorts; 12-week or 12-month window)
7. Acquisition channels in use (organic search, paid social, direct, referral, email, WhatsApp, etc.)

---

What a Cohort Is

A cohort is a group of users who share a defining characteristic within a defined time period. Common cohort definitions:

• All users whose first session occurred in a given week or month (acquisition cohort)
• All users who completed a specific action — made a purchase, downloaded a lead magnet, subscribed to an email list — in a given period (behaviour cohort)

Aggregate metrics (total sessions, total revenue) hide the difference between campaigns that bring one-time buyers and campaigns that build loyal, repeat customers. Cohort analysis reveals which acquisition channels produce high-LTV customers versus one-transaction visitors.

---

Two Cohort Types

Acquisition Cohorts

Users grouped by when they first arrived (Week 1, Week 2, etc.).

Track: What percentage of Week 1 users returned in Week 2, Week 3, Week 4?

Use for: retention analysis, identifying which channels produce loyal audiences.

Behaviour Cohorts

Users grouped by an action they took (first purchase, webinar attendance, lead magnet download).

Track: What percentage converted to the next funnel stage?

Use for: funnel optimisation, identifying where drop-off occurs after a specific action.

---

Building Cohorts in GA4

1. In GA4: Explore → Cohort Exploration
2. Set cohort type:
   • Acquisition date — groups by first session date
   • Event-based — groups by a named event (e.g. purchase, sign_up, generate_lead)
3. Set cohort granularity: weekly (for fast-moving campaigns) or monthly (for longer sales cycles)
4. Set metric: active users, revenue, conversions, or goal completions
5. Set time window: 12-week or 12-month
6. Apply channel filter: segment by Session default channel group to compare organic vs. paid vs. referral vs. WhatsApp cohorts

Permission note: Cohort Exploration requires at minimum Viewer access to GA4. To create custom segments by channel, Editor access is required.

---

Key Insights to Extract

For each cohort analysis, extract and report the following:

Insight	How to read it
Week-4 retention rate	What percentage of Week 1 users are still active 4 weeks later? Under 10% is typical for cold traffic; above 30% indicates a loyal audience
Channel comparison	Which acquisition channel produces the highest Week-4 retention rate?
Revenue by cohort	Which cohort contributes the most total revenue over 6 months?
Decay curve shape	Slow decay = loyal audience building. Steep drop after Week 1 = one-time curiosity traffic — review content and offer alignment

---

Translating Cohort Data for Clients

Do not present raw cohort tables to clients — they cannot interpret colour-coded retention grids without guidance. Translate every cohort analysis into three plain-language client statements:

Statement 1 — Retention: "Of every 100 people who found you through [channel] in [month], [X] were still engaging with your brand 4 weeks later."

Statement 2 — Channel comparison: "Your [channel A] audience retains twice as well as your [channel B] audience — meaning [channel A] produces more durable customers at the same acquisition cost."

Statement 3 — Cohort revenue: "Your [month] cohort is your most valuable — they have generated [X]% more revenue per customer than the [earlier month] cohort."

Pair each statement with a single clear chart (line chart showing retention decay by channel). See meta-dashboard-design for chart selection and mobile-first design rules.

---

Cohort Reporting Output Format

Generate a cohort analysis report structured as follows:

Section 1 — Executive Summary (3 sentences) What the cohort data shows at a glance. Which channel or period is performing best. The single recommended action.

Section 2 — Acquisition Cohort Table Present a simplified cohort table (Week 0 through Week 8 maximum) with the top 3 acquisition channels compared. Highlight the Week-4 retention row.

Section 3 — Behaviour Cohort Funnel If behaviour cohort data is available: show the conversion percentage from acquisition action to next funnel stage for each cohort period.

Section 4 — Channel Comparison Summary A ranked list of channels by Week-4 retention rate. One sentence interpretation per channel.

Section 5 — Recommendations Three SMART actions derived from the cohort data. Format each as:

• Recommendation: [action]
• Rationale: [what the cohort data shows]
• Success metric: [how to measure the outcome]

---

EA-Specific Considerations

• WhatsApp as an acquisition channel: GA4 does not automatically track WhatsApp referrals. Advise the client to use UTM parameters on all WhatsApp links (e.g. ?utm_source=whatsapp&utm_medium=social&utm_campaign=[name]). See meta-utm-tracking for UTM setup.
• Mobile-first data: In Uganda/EA, the majority of sessions originate from mobile devices. Cohort analysis should always segment by device type to identify whether mobile vs. desktop users retain differently.
• Short purchase cycles: For EA e-commerce and service businesses, use weekly cohorts rather than monthly — the typical EA purchase decision cycle is shorter than in Western markets, and monthly cohorts lose resolution.

---

Quality Criteria

Output meets the standard for this skill if:

• Every cohort insight is translated into a plain-language client statement — no raw data tables presented without interpretation
• The report distinguishes between acquisition cohorts and behaviour cohorts and uses the correct type for the client's stated goal
• Channel comparison is included, with at least two channels compared by retention rate
• At least three SMART recommendations are derived directly from the cohort data — not generic analytics advice
• WhatsApp is addressed as an acquisition channel if the client uses it for customer acquisition
• The Week-4 retention rate is calculated and contextualised against the 10% (cold traffic) and 30% (loyal audience) benchmarks
• All monetary values use the client's local currency (UGX for Uganda; KES for Kenya) unless otherwise specified
• Language is British English throughout; imperative in all instructional sections