Analyze · Churn / Retention & Expansion Signals

Analyze Retention Patterns by Client Cohort

Group clients into cohorts by sign-up quarter, channel, or service line and see which ones actually stick.

foundermanagerIntermediate⏱ 1 full day of analyst work
When to use
Use when you suspect certain types of clients churn faster than others but can't prove it. Helpful before changing your ICP, pausing a lead channel, or repricing a service line. Best run quarterly with at least 6 months of cohort data.
The prompt
You are a head of client services at a digital marketing agency running cohort retention analysis to inform ICP and pricing decisions.
Agency: [AGENCY_NAME] — [SERVICES]
Client cohort data (last 24 months):
[COHORT_DATA]
Each row should include: client name, signup month, acquisition channel (referral / outbound / inbound / partner), entry service, entry MRR, status (active / churned / expanded), months retained, churn reason if known.
Group clients into meaningful cohorts and analyze which cohorts retain longest, which churn fastest, and what attributes the high-retention cohorts share. Surface the patterns that should change how the agency sells.

- Build cohorts along at least three axes: signup quarter, acquisition channel, and entry service line
- Calculate average months retained and % still active per cohort
- Call out cohorts with 
Three tables (one per cohort axis) with: Cohort | # Clients | Avg Months Retained | % Still Active | Avg MRR. Then a "What this means" section with 3 ranked recommendations.
Variables
  • [AGENCY_NAME] — Your agency name
  • [SERVICES] — Service lines offered
  • [COHORT_DATA] — CSV-style list: client, signup month, channel, entry service, entry MRR, status, months retained, churn reason
Example input
Agency: Mast & Co — SEO, paid media, web design. 48 clients over 24 months. Sample: Acme Roofing, 2024-Q1, referral, SEO, $6k, active, 16mo; Lyra Beauty, 2024-Q2, outbound, paid media, $4k, churned, 4mo, reason: no leads; Greenfield Law, 2024-Q3, referral, SEO+web, $9k, active, 12mo; [+45 more rows]
Example output
By Channel:
| Channel | # | Avg Months | % Active | Avg MRR |
|---|---|---|---|---|
| Referral | 22 | 14.2 | 73% | $7.4k |
| Outbound | 14 | 5.1 | 21% | $4.2k |
| Inbound | 12 | 9.8 | 58% | $5.9k |

By Entry Service: SEO-only clients retain 13mo avg; paid-media-only retain 5mo (worst). Bundled SEO+web design retain 17mo.

What this means:
1. Outbound clients churn 3x faster at lower MRR — either re-qualify harder or pause the channel
2. Paid-media-only retainers are a churn trap; require a content or SEO attach to sign
3. Referrals over $7k MRR are your gold cohort — build a referral incentive program before doing more outbound
Pro tips
  • Run cohorts by signup quarter to spot whether new sales hires are bringing in worse clients
  • Always cross-cut by entry MRR band — sub-$3k clients almost always retain worse regardless of channel
  • If a cohort has under 8 clients, flag it but don't change strategy on it yet
Works with
ClaudeChatGPTGemini
Done with prompts? Time to install the system
Book a STAOS call
Related prompts