Cohort analysis answers the question that aggregate reporting can't: for customers who first bought in a given week/month, how did they actually behave over the following weeks/months? It's the only way to know your true LTV, your real retention rate, and whether recent changes you've made are building a more loyal customer base or a less loyal one.
How cohorts work in MRP
MRP groups customers by the week/month of their first purchase. Each row in the cohort table represents one acquisition week/month (e.g. June 2024). The columns track how many units those customers bought in subsequent months. Month 0 is the same month they first bought, month 1 is the next month, and so on.
Step-by-step: reading the cohort table
Go to Advanced Tools → LTV & Subscriptions → Cohort Analysis
Start with the Units per Customer view — this is the clearest starting point
Find your most recent complete cohort month (avoid the last 1–2 months which are still filling in)
Read across the row: month 0 shows the average units in the first order. Month 1, 2, 3 show how LTV builds as customers return
Look for the month where growth starts to flatten — this is your typical LTV maturity point and helps you set realistic payback period expectations
Filter to a specific product using the parent ASIN selector below the table to compare LTV across your catalog
Key Cohort views and when to use each
View | When to use it |
Units per Customer | Understand overall LTV maturity. How many orders does a typical customer make within 6–12 months? |
Revenue per Customer | Track how your revenue per customer changes as you adjust pricing over time |
Profit per Customer | See if you're making or losing money in the first month — and when you cross into profit |
Profit Margin | Track margin improvement over a customer's lifecycle; useful when justifying aggressive first-order ACOS to stakeholders |
Retention % | What % of customers make a second purchase within X months? Use to set retargeting windows. |
Using cohorts for testing: When you change your pricing, coupon, or main image, create a mental marker at that week/month. Compare the cohort from that month to cohorts from the 2–3 months prior. If retention and LTV improve for the new cohort, your change worked — for the long term, not just the first order.
S&S vs. non-S&S. Order Type filter
Use the order type filter to compare S&S and non-S&S cohorts separately. For many products, S&S customers have substantially higher LTV. If this is the case for your brand, it justifies more aggressive S&S acquisition (higher coupon, lower price) because the long-term value difference is measurable.