Analytics & Tracking

Cohort Analysis for Shopify: How to Read Retention Curves

How to read Shopify's customer cohort analysis report, what retention curve shapes mean, which cuts to make and how to turn cohorts into CAC payback.

TL;DR
  • A cohort analysis groups customers by when they first bought and follows each group over time. It answers questions a single repeat rate can't, such as whether newer customers come back as well as older ones did.
  • In Shopify's customer cohort analysis report, read across a row to see one cohort age, read down a column to compare cohorts at the same age, and watch where the retention curve flattens. That flat part is your loyal base.
  • Cut cohorts by first product, acquisition channel and first-order discount, then convert cumulative spend into margin to see how long new customers take to pay back their acquisition cost.

What is a cohort analysis, and why use one?

Cohort analysis splits customers into groups that share a starting point, usually the month of their first order, and tracks what each group does afterwards.

A blended metric such as overall repeat purchase rate mixes customers acquired three years ago with customers acquired last week. If acquisition grows quickly, the blended number can fall even when every cohort behaves exactly like the last one, simply because the new customers haven’t had time to come back. Cohorts remove that distortion by comparing customers at the same age.

Cohorts answer practical questions:

  • Are customers acquired this year coming back as often as last year’s?
  • Did the change to the post-purchase flow in spring improve second orders?
  • Do customers from paid social retain worse than customers from search?
  • How many months does a new customer take to repay what we spent acquiring them?

How does Shopify’s customer cohort analysis report work?

Shopify includes a Customer cohort analysis report in Analytics. According to Shopify’s help center, it groups customers into cohorts by the date of their first order. Report availability can depend on your plan, so check what your store has access to.

How the table is laid out:

  • Each row is a cohort: customers whose first order fell in the same week, month or quarter.
  • The first columns name the cohort, show the total for the selected metric and show the value for the cohort’s first orders.
  • The remaining columns show the metric in each period since the first order. Period 0 captures repeat orders placed in the same period as the first order.

You can switch the metric between number of customers, customer retention rate, gross sales, net sales, average order value and amount spent per customer. The default view is a heatmap; you can switch to a retention curve. You can also filter cohorts by first-order attributes such as sales channel, marketing channel or product, filter by customer location or email subscription status, and click a cell for more detail. Shopify only shows spend projections when the metric is amount spent per customer and the store has 24 months of data.

For most stores, the two views worth starting with are customer retention rate by month and amount spent per customer by month.

How do you read a cohort table?

Here is a simplified, illustrative cohort table showing the share of each cohort that placed a repeat order in each month after the first. The numbers are made up to show the method.

First order month Customers Month 1 Month 2 Month 3 Month 6
January 1,000 12% 9% 8% 6%
February 1,200 11% 8% 7% 6%
March 1,500 14% 10% 9% not yet
April 1,400 15% 11% not yet not yet

Read it three ways:

  • Across a row: how one cohort behaves as it ages. January’s activity falls quickly, then settles.
  • Down a column: how different cohorts compare at the same age. Month 1 has improved from 11–12% to 14–15% in March and April. Something changed around March, and you should find out what.
  • The empty cells: recent cohorts haven’t reached later months. Never compare a cohort’s month 6 with another’s month 2.

Two definitions to confirm before you draw conclusions:

  • Retention in a period vs. cumulative retention. “12% in month 1” can mean 12% bought in that month, or 12% have bought at least once more by then. These produce very different curves. Know which one you are reading.
  • Period 0. Repeat orders in the same period as the first order can be significant for stores with fast reorders, and they are easy to overlook.

What do retention curve shapes tell you?

Plot each cohort’s retention by period and the shape of the line tells you more than any single number.

Curve shape What it usually means What to do
Drops sharply, then flattens A loyal core forms after the first few months Find what the flat-line customers have in common and acquire more like them
Keeps falling toward zero Little habit forms; customers buy once and drift away Work on the second order: product fit, onboarding, post-purchase flow
Dips, then rises again later Customers return on a long cycle, or campaigns reactivate them Time reminders to the cycle; check whether a promotion caused the rise
Newer cohorts sit higher Recent changes are working, or the customer mix improved Confirm the cause before scaling it
Newer cohorts sit lower Acquisition is reaching less loyal customers, or the experience slipped Cut by channel and first product to find the source

The flat part of the curve matters most. It shows the share of customers who become regulars, and it is the base that customer lifetime value builds on.

Which cohort cuts are worth making?

The default monthly view tells you whether something changed. Cuts tell you why.

By first product

Customers who start with some products come back far more often than customers who start with others. Filter by first-order product, or export the data and group by first product. Products with high first-to-second order rate deserve more acquisition budget, even when their first-order margin is lower.

By acquisition channel

Filter by the marketing channel of the first order. If paid social cohorts retain far worse than search or email cohorts, a channel’s first-order ROAS overstates its long-term value, or the reverse.

By first-order discount

Compare customers whose first order used a discount code with those who paid full price. If discount-led cohorts rarely return, your welcome offer may be buying one-time customers.

By season

Black Friday and holiday cohorts often behave differently from the rest of the year. Compare them with each other year over year, not with a March cohort.

Keep cohort size in mind with every cut. A cohort of a few dozen customers moves several points on a handful of orders, so treat small cuts as hints, not answers.

Worked example: from cohort spend to CAC payback

All numbers are illustrative.

A store’s March cohort has 1,000 new customers. The marketing spend that acquired them gives a customer acquisition cost of $45. The report’s amount spent per customer, cumulative, reads:

Months since first order Cumulative spend per customer Cumulative contribution margin at 35%
0 (first order) $70 $24.50
3 $95 $33.25
6 $120 $42.00
9 $135 $47.25
12 $150 $52.50

Steps:

  1. Convert spend to margin. Revenue doesn’t repay acquisition cost; contribution margin does. Multiply cumulative spend by your contribution margin, here 35%. Use net sales if returns are significant.
  2. Find where cumulative margin passes CAC. $42.00 at month 6 is short of $45; $47.25 at month 9 passes it. The CAC payback period for this cohort is between 6 and 9 months.
  3. Compare across cohorts. If April’s cohort pays back in 5 months and May’s needs 11, look at what changed in acquisition between them.
  4. Decide. A payback period that fits your cash position means you can grow spend. One that doesn’t means you need better retention, higher margin or cheaper acquisition before scaling.

What mistakes make cohort analysis misleading?

  • Comparing immature cohorts with mature ones. Only compare at equal ages.
  • Reading revenue as profit. Discount-heavy cohorts can show high spend with thin margin.
  • Ignoring returns. Gross sales overstate value for categories with frequent returns. Use net sales where you can.
  • Mixing customer identities. Repeat orders only connect when Shopify can tie them to the same customer record. Orders placed under different emails look like new customers.
  • Overreacting to one cohort. A single unusual month is often a promotion or a stock issue. Look for a pattern across several cohorts.

Checklist: a monthly cohort review

  • [ ] Open customer retention rate by month and note whether the newest mature cohorts sit above or below earlier ones.
  • [ ] Confirm whether you are reading per-period or cumulative retention.
  • [ ] Switch to retention curves and check where they flatten.
  • [ ] Cut by first product, marketing channel and first-order discount.
  • [ ] Export amount spent per customer and convert it to contribution margin.
  • [ ] Calculate CAC payback for the last three mature cohorts.
  • [ ] Note promotions, launches and stock issues next to the cohorts they affected.
  • [ ] Write down one change to test and the cohort you will watch to judge it.

FAQ

How many months of data do I need for a cohort analysis?

You can start with a few months, but early periods only show short-term behavior. For products bought every month or two, six months shows a useful curve. For slower categories, a year or more gives a fairer picture.

What is the difference between cohort retention and customer retention rate?

Customer retention rate usually measures your whole customer base over one period. Cohort retention follows one starting group over time. Our guide to calculating customer retention rate covers the period-based formula.

Should I use weekly, monthly or quarterly cohorts?

Monthly suits most stores. Weekly cohorts are useful for fast-moving consumables with high order volume. Quarterly cohorts help low-volume or high-ticket stores where monthly groups are too small to read.

Turn this into a plan for your store

The Free Profit Audit scans your store, sizes your biggest profit leaks and builds a 90-day plan.

Free Profit Audit