Analyze Customer Cohort Retention

You are a growth analyst for a digital marketing agency analyzing the agency's own GTM ROI. You translate cohort tables into retention curves and root-cause hypotheses.
Agency: [AGENCY_NAME] — [SERVICES] | Period: [PERIOD] | Data:
[COHORT_DATA] (rows = signup month, columns = M0, M1, M2 ... active client counts)
Calculate retention % at M1, M3, M6, and M12 for every cohort. Identify the strongest and weakest cohorts, quantify the gap, and hypothesize 2-3 reasons grounded in the numbers (intake channel, pricing tier, onboarding change).

- Show the formula: Retention(Mn) = Active(Mn) / Active(M0) × 100
- Flag any cohort with 15pp as a churn cliff
- Recommend one concrete experiment (offer, onboarding step, ICP filter) to lift the weakest cohort

1) Retention matrix table (cohort × M1/M3/M6/M12 with %). 2) "Best vs Worst" 2-row comparison. 3) Hypotheses (bulleted, each tied to a number). 4) One recommended experiment with expected lift.

• [AGENCY_NAME] — Your agency name
• [SERVICES] — Primary services sold (e.g., SEO + paid social retainers)
• [PERIOD] — Analysis window (e.g., Jan 2025 - Apr 2026)
• [COHORT_DATA] — Table of signup-month cohorts and active client counts by month-since-signup

Agency: Northbeam Digital — SEO + paid search retainers | Period: Jan 2025 - Apr 2026 | Data:
Cohort | M0 | M1 | M3 | M6 | M12
Jan 2025 | 12 | 12 | 11 | 9 | 6
Apr 2025 | 9 | 9 | 8 | 5 | 3
Jul 2025 | 14 | 13 | 11 | 10 | -
Oct 2025 | 11 | 11 | 9 | - | -
Jan 2026 | 8 | 8 | - | - | -

Retention matrix:
Jan'25: M1 100% / M3 92% / M6 75% / M12 50%
Apr'25: M1 100% / M3 89% / M6 56% / M12 33%
Jul'25: M1 93% / M3 79% / M6 71% / M12 early signal
Oct'25: M1 100% / M3 82% / M6 early signal
Jan'26: early signal only

Best vs Worst: Jan'25 (50% M12) vs Apr'25 (33% M12) — 17pp gap.

Churn cliff: Apr'25 lost 33pp between M3 and M6 — a contract-renewal cliff.

Hypotheses:
- Apr'25 cohort signed during the Q2 "$1.5k SEO starter" promo — likely under-priced clients with lower commitment.
- Jul'25 onboarding overhaul (kickoff scorecard) appears to be holding M3 retention at 79% despite higher intake volume.
- Sub-$2k MRR accounts churn at 2x the rate of $3k+.

Experiment: Sunset the starter tier; require $2.5k MRR minimum on new contracts. Expected lift: +10-15pp M6 retention based on Jan'25 vs Apr'25 delta.

• Run this the same week each quarter so cohort ages line up consistently.
• Layer in MRR per cohort, not just headcount — a 50% retained cohort at $5k MRR beats 80% retained at $1k.
• Tag each cohort with its dominant intake channel so the next prompt (LTV by channel) can build on this output.