Cohort Economics and Unit Profitability: Understanding Cohort Performance
Master cohort analysis. Analyze profitability, identify trends, optimize segments.
Key Takeaways
- Cohort definition: Group of customers acquired in same time period (month, quarter). Analysis: Track cohort over time (does Q1 2024 cohort stay, spend, expand?). Metrics: Retention curve (% retained over time), revenue per cohort member, lifetime value by cohort. Purpose: (1) Understand product quality (better product = better retention cohorts), (2) Identify acquisition quality (expensive customer source = slower payback), (3) Measure improvements (new onboarding better retention).
- Cohort economics: (1) CAC for cohort (how much spent to acquire), (2) Revenue per customer (ARPU × lifetime), (3) Profitability (revenue - CAC). Compare cohorts: Q1 cohort £1500 LTV, Q2 £1700 LTV (20% improvement = better product or cheaper CAC). Identify trends: Improving cohorts = product/retention getting better. Declining = problem (churn worsening, or expensive acquisition).
- Analysis example: Q1 cohort (100 customers, £500 CAC each). Month 1 revenue: £100 × £200 = £20K. Month 3: 85 customers (15% churn) = £17K. Month 6: 70 customers (30% total churn) = £14K. Month 12: 54 customers (46% churn) = £10.8K. LTV: (£20K + £17K + £14K + ... over 12 months) / 100 = ~£150K total / £50K CAC = £100K net LTV. Profitable if revenue exceeds CAC quickly enough.
Analyzing Cohort Economics and Profitability
Understanding unit profitability through cohort-level analysis. **Cohort analysis fundamentals** Definition: - Cohort: Group of customers acquired in same period (month, quarter, year) - Analysis: Track cohort's revenue, retention, spending over time - Purpose: Understand customer quality and profitability by acquisition period Why cohorts matter: Example: Two cohorts, same £500 CAC, different results Cohort Q1 (jan-mar 2024): - Acquisition CAC: £500 per customer - Month 1 retention: 95% - Month 6 retention: 80% - Month 12 retention: 60% Cohort Q2 (apr-jun 2024): - Acquisition CAC: £500 per customer - Month 1 retention: 92% - Month 6 retention: 75% - Month 12 retention: 50% Both cost £500, but Q1 retains 60% vs Q2 50% = Q1 is 20% more valuable Insight: Q1 acquisition source better (higher-quality customers) OR onboarding improved (Q1 got better training) **Cohort retention curves** Definition: - Visualization of cohort retention over time - X-axis: Months since activation - Y-axis: % of cohort retained Example cohort table: | Cohort | Month 0 | Month 1 | Month 3 | Month 6 | Month 12 | Month 24 | |---|---|---|---|---|---|---| | Jan 2023 | 100 | 95 | 85 | 72 | 54 | 38 | | Feb 2023 | 100 | 96 | 87 | 75 | 58 | 42 | | Mar 2023 | 100 | 97 | 89 | 78 | 62 | 45 | Interpretation: - Retention improving (Mar cohort better than Jan) - Jan cohort: Lose 46% in 12 months, 62% in 24 months - Mar cohort: Lose 38% in 12 months, 55% in 24 months - Product or onboarding improving (later cohorts retain better) **Cohort revenue analysis** Definition: - Revenue per cohort member over time - Tracks: ARPU growth, expansion, downgrades Example: 100 customers per cohort, £100/month starting price | Cohort | Month 1 | Month 3 | Month 6 | Month 12 | Total revenue | |---|---|---|---|---|---| | Jan 2024 | £10K | £8.5K | £7.2K | £5.4K | £102K per 100 | | Feb 2024 | £10K | £8.7K | £7.5K | £5.8K | £106K per 100 | | Mar 2024 | £10K | £8.9K | £7.8K | £6.2K | £111K per 100 | Interpretation: - Revenue declining over time (churn = fewer customers) - Later cohorts generating more revenue (better retention) - Total 12-month revenue: Jan £1.02K per customer, Mar £1.11K per customer (9% higher) **Cohort profitability analysis** Formula: Cohort profitability = Total revenue - CAC - COGS Example: Jan 2024 cohort | Item | Amount | |---|---| | 100 customers acquired | 100 | | CAC per customer | £500 | | Total acquisition cost | £50K | | 12-month revenue per cohort | £102K (from above) | | COGS (30% of revenue) | £30.6K | | Gross profit | £71.4K | | Acquisition cost | -£50K | | Net profit | £21.4K | | Profit per customer | £214 | | Payback period | 7.1 months | Calculation detail: - First month: Revenue £10K, COGS £3K, gross profit £7K - Acquisition cost was £50K (all upfront) - Profit per month: £10K × 70% = £7K - Payback: £50K / £7K = 7.1 months - After month 7.1, customer is profitable (every month adds profit) Comparison across cohorts: | Cohort | Revenue | COGS | Gross | CAC | Net profit | Payback | |---|---|---|---|---|---|---| | Q1 2023 | £102K | £30.6K | £71.4K | £50K | £21.4K | 7.1 months | | Q2 2023 | £106K | £31.8K | £74.2K | £50K | £24.2K | 6.8 months | | Q3 2023 | £111K | £33.3K | £77.7K | £50K | £27.7K | 6.4 months | Insight: Later cohorts more profitable (faster payback, higher revenue) Reasons: - Better product (fewer bugs = less support costs = lower COGS) - Better onboarding (customers adopt faster = lower early churn = higher revenue) - Better CAC (more efficient acquisition = lower cost) **Cohort economics by segment** Different segments, different economics: Enterprise cohort (£1000 CAC, £5000 ARPU): - 100 customers × £1000 CAC = £100K acquisition - Month 1 retention: 98% - Month 12 retention: 88% - 12-month revenue: £5K × 10 months average = £50K per customer - Total revenue: £5M - COGS (20% for enterprise): £1M - Gross profit: £4M - Net profit: £4M - £100K = £3.9M - Payback: 2.4 months (very fast) - Profit per customer: £39K (excellent) SMB cohort (£300 CAC, £50 ARPU): - 1000 customers × £300 CAC = £300K acquisition - Month 1 retention: 80% - Month 12 retention: 40% - 12-month revenue: £50 × 6 months average = £300 per customer - Total revenue: £300K - COGS (40% for SMB): £120K - Gross profit: £180K - Net profit: £180K - £300K = -£120K (UNPROFITABLE!) - Payback: Never (loses money) - Profit per customer: -£120 (negative) Insight: Enterprise very profitable, SMB not. Strategy: - Invest in enterprise (high CAC justified by LTV) - Reduce SMB CAC (move to self-serve, reduce acquisition cost) OR - Increase SMB LTV (reduce churn, improve retention) OR - Exit SMB (focus on higher-margin) **Using cohort analysis for optimization** Scenario 1: Declining retention cohorts Observation: Jan 2024 cohort 60% retention, Feb 65%, Mar 70% (improving trend) Hypothesis: Onboarding improvement in March Actions: - Measure: What changed in March? (new onboarding process? CS team hired?) - Replicate: Apply March process to all customers - Expected: Future cohorts all retain 70% (instead of varying 60-70%) Impact: If average prior retention 65%, improve to 70% = 7.7% retention improvement Revenue impact (1000 customer base): - 12-month revenue at 65% retention: 1000 × £100 × 7 months average = £700K - 12-month revenue at 70% retention: 1000 × £100 × 7.5 months average = £750K - Improvement: £50K (7% revenue increase) Scenario 2: Expensive cohort (slow payback) Observation: Q1 cohort 7.1 month payback, Q2 6.8 months, Q3 6.4 months (improving) But Q1 acquisition source still expensive. Payback still slow vs target 4 months. Actions: - Identify Q1 source (which channel cost £500 CAC?) - Compare to Q3 (which achieved £500 CAC with 6.4 month payback) - Shift budget to Q3 source (more efficient) - Kill Q1 channel (too expensive) Expected: Shift acquisition budget to efficient channels, improve payback to 5-6 months Scenario 3: Product quality signal Observation: Retention curves improving over time Jan cohort: 60% month 12 Feb cohort: 62% month 12 Mar cohort: 65% month 12 Apr cohort: 68% month 12 (projected) Insight: Product quality improving (product updates, features, bug fixes) Validate: Check feature releases, customer feedback, churn reason data Confidence: Cohort trend is strong signal (consistent improvement) **Cohort dashboard** Monthly monitoring: | Cohort | Size | Retention | Revenue/customer | LTV | CAC | Net LTV | Status | |---|---|---|---|---|---|---|---| | Jan 24 | 50 | 60% | £1000 | £1500 | £500 | £1000 | Good | | Feb 24 | 60 | 62% | £1050 | £1575 | £500 | £1075 | Better | | Mar 24 | 70 | 65% | £1100 | £1650 | £500 | £1150 | Better | | Apr 24 | 75 | 67% | £1120 | £1680 | £520 | £1160 | Better | Metrics to watch: - Retention trend (improving, flat, declining?) - LTV trend (getting better, worse?) - CAC stability (staying consistent?) - Net LTV (profit per customer) Quarterly review: - Compare to prior quarter (are cohorts getting better?) - Identify inflection points (when did something change?) - Correlate with product/operational changes (what drove improvement?) **Common cohort analysis mistakes** Mistake 1: No cohort analysis (aggregate only) - Problem: Know overall 5% churn, don't know if cohorts differ - Fix: Segment by cohort, compare retention curves - Impact: Identify trends, product improvements Mistake 2: Analyze retention without revenue - Problem: Track retention, miss that some cohorts have lower ARPU - Fix: Combine retention + revenue per cohort - Impact: Understand true profitability Mistake 3: No connection to product/operations - Problem: See cohort trends, don't investigate root cause - Fix: Correlate cohort trends with product changes, acquisitions changes - Impact: Validate product improvements work, identify what drove changes Mistake 4: Ignore expensive cohorts - Problem: Some cohorts have 10-month payback (slow), don't act - Fix: Kill expensive acquisition sources, shift budget to efficient - Impact: Improve unit economics, payback