Cohort Analysis: The Secret to Understanding Your Store's Real Health
Your monthly revenue is up 20%. Congratulations? Maybe not. Revenue growth can mask serious problems—declining retention, weakening cohorts, unsustainable acquisition spending. The only way to truly understand your store's health is cohort analysis: tracking groups of customers over time to see if they're getting more valuable or less valuable. It's the difference between looking at a dashboard and actually understanding your business.
Why Monthly Revenue Metrics Lie
Let's start with why the metrics most stores obsess over—monthly revenue, monthly orders, monthly growth—are fundamentally misleading:
The Dangerous Scenario
January Revenue: $100,000 (500 new customers, $200 AOV)
February Revenue: $120,000 (600 new customers, $200 AOV)
March Revenue: $144,000 (720 new customers, $200 AOV)
Growth Rate: 20% month-over-month. Amazing, right?
The hidden truth: January customers had a 30% repurchase rate. February customers had 22%. March customers had 18%. Your retention is collapsing while revenue grows. You're building a house of cards fueled by unsustainable acquisition spending.
Monthly revenue only tells you what happened this month. It doesn't tell you if your business is getting healthier or sicker. That requires cohort analysis.
Key Insight
A cohort is a group of customers who share a common characteristic—typically, the month they made their first purchase. By tracking cohorts over time, you see if your business model is improving (newer cohorts perform better) or deteriorating (newer cohorts perform worse). This is your true health indicator.
What Cohort Analysis Actually Reveals
When you analyze customers by cohort, patterns emerge that aggregate metrics completely miss. Here's what you can see:
1. Retention Curve by Cohort
Track what percentage of each cohort makes a second purchase, third purchase, fourth purchase. This reveals your retention stability.
Example Cohort Retention:
Jan Cohort: Month 1 (30%), Month 3 (22%), Month 6 (18%)
Feb Cohort: Month 1 (28%), Month 3 (20%), Month 6 (15%)
Mar Cohort: Month 1 (25%), Month 3 (17%), Month 6 (??)
Pattern: Each cohort is retaining worse than the previous one. You have a retention problem that's getting worse.
2. Revenue Curve by Cohort
Track cumulative revenue per customer for each cohort over time. This shows if newer customers are more or less valuable than older ones.
Example Revenue Progression:
Jan Cohort: Month 0 ($75), Month 3 ($140), Month 6 ($220), Month 12 ($350)
Feb Cohort: Month 0 ($75), Month 3 ($125), Month 6 ($185), Month 12 (projected $280)
Mar Cohort: Month 0 ($75), Month 3 ($110), Month 6 (projected $160)
Pattern: January customers generated $350 LTV. March customers projected $240 LTV. Customer quality is declining 32%.
3. Payback Period by Cohort
If you know acquisition cost per cohort, you can calculate how long it takes to break even. This tells you if you're spending sustainably.
Example Payback Analysis:
Jan Cohort: CAC $60, Month 0 profit $25, Month 3 cumulative $65 (broke even at Month 3)
Feb Cohort: CAC $70, Month 0 profit $25, Month 3 cumulative $55 (broke even at Month 5)
Mar Cohort: CAC $85, Month 0 profit $25, Month 3 cumulative $45 (won't break even until Month 8+)
Pattern: Your payback period is expanding. If this continues, you'll run out of cash before customers become profitable.
4. Channel Performance by Cohort
Not all acquisition channels produce equal cohorts. Breaking down cohorts by channel reveals which sources bring high-value customers.
Example Channel Cohorts (6-Month LTV):
Organic cohort: $285 LTV, 35% retention
Google Ads cohort: $215 LTV, 28% retention
Facebook Ads cohort: $165 LTV, 19% retention
Insight: You're spending equally across channels, but organic delivers 72% more LTV than Facebook. Reallocate budget immediately.
How to Run Cohort Analysis (Step by Step)
You don't need fancy tools to start cohort analysis. Here's how to do it manually:
Step 1: Export Your Order Data
Export all orders from your store with: customer ID, order date, order value, product details, acquisition channel. Go back at least 12 months if possible.
Step 2: Identify First Purchase Date for Each Customer
For each customer, find their first order date. This determines their cohort. A customer who first bought in January 2024 belongs to the "Jan 2024 cohort."
Step 3: Calculate Time Since First Purchase
For every order, calculate "months since first purchase." This lets you track cohort behavior over time. An order 3 months after first purchase gets labeled "Month 3."
Step 4: Create a Cohort Table
Build a table with cohorts as rows and time periods as columns:
Cohort | M0 | M1 | M2 | M3 | M6 | M12 Jan 2024 | $75k | $22k | $18k | $15k | $12k | $10k Feb 2024 | $80k | $20k | $16k | $13k | $10k | ... Mar 2024 | $85k | $19k | $15k | $11k | ... | ...
Step 5: Analyze Patterns
Look for trends. Are newer cohorts performing better or worse? Is retention improving or declining? Are revenue curves steeper or flatter? These patterns tell your store's real story.
What Good vs. Bad Cohort Patterns Look Like
Here's how to interpret your cohort analysis results:
Healthy Cohort Pattern
- Newer cohorts retain at same or better rates than older ones
- Revenue curves are stable or improving over time
- Payback period is stable or decreasing
- Cohorts become profitable within 3-6 months
What this means: Your business model is working. You're acquiring better customers or retaining them better. This is sustainable growth.
Unhealthy Cohort Pattern
- Newer cohorts retain worse than older ones
- Revenue curves are declining or flattening
- Payback period is expanding
- Recent cohorts won't break even for 12+ months
What this means: Your business model is deteriorating. Customer quality is declining, or retention is weakening. This is unsustainable—revenue growth is masking a serious problem.
How to Fix Declining Cohort Performance
If your cohort analysis reveals declining performance, here's your action plan:
1. Segment by Acquisition Channel
Run cohort analysis separately for each channel. Find which channels are driving the decline and cut or optimize them. Double down on channels with strong cohorts.
2. Analyze Product Mix Changes
If newer cohorts are buying different products than older cohorts, that might explain performance changes. Products with lower repeat rates drag down cohort performance.
3. Improve First-Purchase Experience
If newer cohorts have worse retention from Month 1, the problem is likely in onboarding, product quality, or fulfillment. Fix the first-purchase experience to improve cohort retention curves.
4. Implement Cohort-Specific Retention Campaigns
Build targeted campaigns based on cohort age. Month 1 cohorts need onboarding and education. Month 3 cohorts need replenishment reminders. Month 6+ cohorts need win-back or upgrade offers.
5. Reduce Discount-Driven Acquisition
If you've increased promotional intensity, you might be attracting lower-quality customers who only buy on discount. Cohort analysis will show this as declining LTV curves. Pull back on discounts and focus on value-based acquisition.
How Lumino Automates Cohort Analysis
Manual cohort analysis is time-consuming and error-prone. Lumino does it automatically—tracking every cohort, calculating retention curves, projecting LTV, and comparing performance across channels, products, and time periods. You get instant visibility into your store's true health without spreadsheets or manual calculations.
But Lumino doesn't just show you cohort data—it interprets it. It tells you which cohorts are underperforming, why, and what to do about it. You get intelligence, not just charts.
The Bottom Line
Monthly revenue can lie. Monthly growth can mask decay. The only way to understand your store's true health is cohort analysis—tracking groups of customers over time to see if your business model is improving or deteriorating. If you're not doing cohort analysis, you're flying blind. And in e-commerce, that's a recipe for running out of runway before you realize there's a problem.