Cohort analysis: Read ecommerce retention

Cohort analysis follows a group of customers over time, so you can see whether they return and buy again. For ecommerce, start with customers who made their first purchase in the same month. Compare groups at the same age before deciding whether retention is improving.

The main trap is comparing different measures. Customers buying during a particular month and customers who have bought again by a deadline answer different questions.

What is cohort analysis?

A cohort is a group with a shared starting point or experience. Cohort analysis tracks what happens to that group afterward, rather than combining new and established customers.

An acquisition cohort might contain customers whose first purchase happened during a holiday sale. A behavioral cohort could contain customers who bought the same starter product. Following those groups can reveal whether a successful acquisition campaign also brings customers who return.

Total sales can rise while newer customers become less likely to purchase again. A cohort view helps you notice that change before treating growing revenue as proof that retention is healthy. It adds detail to your broader customer retention strategy.

A segment is a customer group you can describe or target. A cohort analysis adds the time dimension: what happened to a defined group after its shared start? Keep its membership consistent when comparing later results.

How do you read a cohort table?

Read across a row to follow one group, and down a column to compare groups at the same age. Check the metric label before interpreting the colors.

• Rows identify the group. A row might contain customers who first bought in January.
• Columns show elapsed time. They describe how long it has been since the first purchase, rather than a shared calendar month.
• Cells show the outcome. Depending on the report, this could be customers buying during that period or customers who have bought again by then.

Imagine a customer first buys in January, buys again in February, and places no order in March. In a monthly activity table, that customer appears among February's returning buyers but not March's. In a cumulative second-purchase view, that customer remains someone who has bought again after February.

This distinction changes how you read the shape. Monthly buying activity can rise and fall. A cumulative share answering “have they bought again yet?” can only rise or stay unchanged as time passes.

If several groups weaken during the same calendar period, investigate what affected customers then, such as product availability. If only one acquisition group weakens, examine its offer and customer mix. These patterns suggest questions to investigate; they do not establish a cause.

A blank cell usually means a newer cohort has not reached that age. Treat it as unavailable, rather than as a failure to return. Comparing an older group's later results with a younger group's early results gives the older group more opportunity.

Check the cohort size, too. A small group can show a dramatic percentage change after a few additional purchases. Look for a repeated pattern before changing your offer, and compare holiday cohorts with similar seasonal conditions.

What changes when you give customers longer to return?

The observation window changes what you can conclude about repeat purchasing. A customer who has not bought again soon after delivery may still return later.

These historical benchmarks do not show whether your recent monthly cohorts are improving. Use your own equally mature groups to answer that question.

This benchmark provides context for choosing a comparison window. It does not identify the best day to send an email or prove that waiting longer improves retention. Your own customers' purchase needs should guide the follow-up you test.

It also stops at the second purchase. A customer counted here may or may not place later orders. For the distinction between this measure and a period's repeat buyers, see the repeat purchase rate benchmarks.

Why does your product category matter?

Different products create different reasons to return. Someone replenishing supplements may have a different purchase cycle from someone furnishing a room.

Health food and supplements show a higher second-purchase share than decor and furniture at this deadline. That is a reason to choose relevant context, rather than apply one target to different businesses.

The comparison does not explain why the categories differ. Product range, customer mix, and other differences could matter. Use it to frame a question about your store, rather than justify changing your business to match another category.

How do you turn a cohort comparison into action?

Start with a question that leads to a decision you can test. “Are customers from our sale coming back?” is more useful than reviewing a heatmap without a purpose.

1. Choose the groups. Compare first buyers from the sale with first buyers from a comparable period. Note differences in season, first product, or acquisition channel.
2. Name the outcome. Decide whether you want to see purchases during a later period or a second purchase by a deadline. Keep that definition consistent.
3. Compare at the same age. Give both groups the full observation window. Set aside results they have not had time to reach.
4. Investigate the difference. Review the offer, delivery experience, product availability, and messages those customers received. The chart tells you where to look; customer feedback can help explain why.
5. Test a focused change. Try clearer product guidance or a relevant replenishment reminder, then assess the same outcome. A controlled comparison gives stronger evidence than a before-and-after change alone.

For example, sale buyers might return less often because they bought gifts rather than products for themselves. Sending more reminders may miss that distinction. Check what they bought and what they tell you before assuming the email sequence caused the gap.

Track customer value alongside purchase behavior when the decision concerns acquisition spending. A group can generate more revenue through larger purchases while fewer people return. Revenue, purchase frequency, and customer retention describe different parts of the result.

Keep a short record of the group, outcome, finding, and next action. That makes the next comparison easier to interpret. For practical messaging ideas, use the guide to encouraging repeat purchases.

Which tool should you start with?

Start with a report you already have that shows first-purchase groups and elapsed time. Confirm what its return metric means before comparing it with another report.

A spreadsheet can work for a manageable, occasional comparison. Choose a reporting tool when you need to repeat the analysis or compare several groups regularly. Look for clear metric labels, visible cohort sizes, and an explanation of blank cells.

By the Numbers offers cohort analysis reports, including Cohort Retention and Cohort Repurchase Rate. The former shows buying in later periods; the latter shows buying again within a deadline. Choose the report that matches the question you wrote down.

Methodology

These benchmarks use anonymized, aggregated data from Shopify stores in the By the Numbers network. No individual store is identifiable from any figure. Stores that use an analytics app are not a random sample of all Shopify stores, so read the figures as a reference point, not as an average for the whole market.

Keep reading

If you want the full picture on cohort analysis and retention, start with how to read your cohort retention curve.

Related reading: Repeat Purchase Rate by Industry, Customer Churn Rate on Shopify: Calculate and Reduce It and How to Predict Churn Using Segmentation (Shopify).

By the Numbers builds this into the dashboard. See cohort analysis.