How to use Cohort Revenue Analysis
Cohort revenue analysis groups customers by when (or how) you acquired them and tracks what each group spends over time. It's the fastest way to spot acquisition-quality drift that sitewide LTV charts hide.
Cohort Revenue Analysis
Tracking revenue from groups of customers acquired in the same window to see how their lifetime value evolves.
Cohort revenue analysis groups customers by a shared acquisition event — usually the month they placed their first order, but it can also be the campaign, channel, or landing page that brought them in — and tracks how much each group spends in the weeks and months that follow. Plotted side by side, the cohorts show whether last quarter's new customers are pacing ahead of, behind, or in line with the ones before them.
It sits inside revenue intelligence as the diagnostic view: blended LTV averages mask the fact that a recent paid-social push may be importing buyers who never come back, while an older organic cohort quietly compounds. Cohorts make that trade-off visible.
Most stores look at revenue the wrong way round. They watch sitewide totals — this week vs last week, this month vs last month — and conclude the business is healthy because the line goes up. That line can keep going up while the underlying customer quality is falling apart.
The cohort view fixes that. Instead of asking "how much did we make in March?", you ask "how much have customers we acquired in March spent so far, and how does that compare to customers we acquired in February, January, December?" When those curves start diverging, you have an early warning that sitewide numbers will hide for another two quarters.
What a cohort actually measures
A cohort is just a labelled bucket of customers. The label is the moment they entered your business: their first-order month, the campaign UTM on that first order, the channel (paid social, organic, email, affiliate), or sometimes the discount they used. Pick one labelling scheme per report and stick with it.
The metric you stack on top is usually cumulative revenue per customer at month 1, 2, 3, and so on after acquisition. Some teams use orders-per-customer or repeat rate instead, but revenue per customer is the cleanest because it folds AOV and repeat behaviour into one number you can compare.
The key trick is the relative time axis. Every cohort starts at month zero on the same vertical line, regardless of the calendar month they were acquired. That's what lets you compare a cohort that's six months old against one that's three months old without the chart turning into noise.
Don't compare cohorts that haven't matured
A cohort acquired three weeks ago will always look weaker than one acquired a year ago — it hasn't had time to place repeat orders. Compare like windows: month-3 revenue of the new cohort vs month-3 revenue of older cohorts. Comparing lifetime totals across mismatched ages is the single most common mistake teams make on their first cohort report.
Building the view on your store
On Shopify and WooCommerce the raw data is already there — every order has a customer ID, a created_at timestamp, a total, and (if your tagging is clean) a source attribution. The work is in the pivot. You assign each customer to a cohort based on their first order, then sum revenue by cohort and by months-since-acquisition.
Three practical decisions to make up front: do you count gross or net revenue (after refunds and discounts)? do you include shipping? and do you cohort on first-order date or first-touch date? Most teams land on net revenue excluding shipping, cohorted on first-order date — because that's what actually hits margin.
Cumulative revenue per customer, by acquisition month
Jan cohort
Apr cohort
Jul cohort (paid-social heavy)
Read the chart this way: at month 3, the January and April cohorts are sitting around €102-€104 per customer. The July cohort — acquired during a heavier paid-social push — has only made it to €78 by the same age. That's a 24% drop in three-month customer value, and it would be completely invisible in a sitewide LTV report for another six months.
Reading the numbers by channel
The acquisition-month cut tells you when quality shifted. The acquisition-channel cut tells you why. Splitting the same cohort by where the customer came from usually reveals one or two channels carrying the average and one or two dragging it down.
The pattern below is typical for an apparel store in the €1M-€10M band: organic search and email-acquired customers convert at a higher first-order AOV and repeat meaningfully, while paid-social and discount-driven cohorts spike on day one and flatten.
Typical 6-month revenue per customer by acquisition channel (apparel, €60 AOV)
| Acquisition channel | First-order AOV | 6-mo revenue/customer | Repeat rate (6mo) |
|---|---|---|---|
| Organic search | €68 | €152 | 44% |
| Email / SMS list | €72 | €168 | 51% |
| Paid search (brand) | €66 | €144 | 41% |
| Paid search (non-brand) | €61 | €108 | 28% |
| Paid social (prospecting) | €54 | €86 | 19% |
| Affiliate / discount sites | €49 | €72 | 14% |
What you do with this table matters more than what's in it. If paid-social prospecting is delivering customers worth €86 over six months and your fully-loaded CAC on that channel is €45, you're profitable but thin. If CAC rises to €60 — which happens every Q4 — that channel is now losing money on a six-month horizon, and the cohort view will catch it before the P&L does.
Turning the report into decisions
A cohort report is only useful if it changes what you spend on next month. Three decisions it should drive: reallocating paid budget away from cohorts whose 90-day revenue doesn't clear CAC, doubling down on the channels whose cohorts compound past month 6, and triggering a retention push (flow, win-back, bundle) on underperforming cohorts before they go fully dormant.
The cadence most teams settle on is a monthly cohort review the week the previous month closes. You're looking at the new cohort's month-0 number against prior cohorts' month-0, and the still-maturing cohorts' month-3 and month-6 progress. Anything that's lagging the trend by more than 15% gets escalated.
Pair it with the rest of the revenue intelligence stack
Cohort revenue analysis tells you which groups of customers are worth what. To act on it, you need attribution clean enough to trust the channel labels, retention reporting granular enough to see where in the lifecycle each cohort drops off, and a contribution-margin view so you know which customers are profitable, not just high-revenue.
Frequently asked questions
LTV is usually a single blended number across all your customers. Cohort revenue analysis breaks that average into the groups that produced it, so you can see whether recent acquisitions are above or below the historical curve. LTV tells you where you've been; cohorts tell you where you're going.
Around 200-300 customers per cohort is where the curves stabilise for most DTC stores. Below that, a handful of high-spending or returning customers can swing the average enough to mislead you. If you're smaller, widen the cohort window from monthly to quarterly.
First-order date is more reliable because the data is unambiguous and lives in your store database. First-touch date depends on attribution that may break across devices, ad blockers, and iOS privacy changes. Use first-order date as your default; reserve first-touch cohorts for specific campaign post-mortems.
Shopify's native reports include a basic cohort view under Analytics, but it's limited to monthly cohorts and doesn't let you split by channel or campaign. For anything past a quick check you'll want to export orders to a BI tool, a Sheets pivot, or a platform that builds the cohorts for you.
At least 12 months, ideally 24. You need enough cohort history to see what a mature cohort curve looks like for your business — that's your benchmark for whether new cohorts are tracking ahead or behind. Importing historical GA4 and order data on day one removes the cold-start problem.
Net of refunds and discounts, excluding shipping, is the standard. It approximates what actually hits your bank account and means a cohort that bought heavily on a 30%-off code doesn't look artificially strong. If you discount heavily, this distinction is the difference between a cohort looking healthy and looking dangerous.
Keep them in the same report but tag them as a separate cohort line. Subscription cohorts have radically different shapes — steady monthly revenue rather than spiky repeat purchases — and mixing them into a one-time-purchase cohort skews the average. Most teams maintain two parallel cohort views: subscribers and non-subscribers.
For apparel, beauty, and home goods in the €40-€80 AOV range, healthy cohorts hit 2.0x-2.5x first-order AOV by month 6. Consumables and supplements run higher, often 3x+. Single-purchase categories like furniture obviously sit closer to 1.1x-1.3x. Compare your cohorts to your own historical curve before benchmarking externally.
Yes, and it's one of the most underused versions of the report. Tag every customer with the campaign that acquired them and you can run a cohort curve per launch, per influencer drop, per Black Friday window. It tells you which campaigns brought in customers worth keeping and which just bought one-time discount hunters.
You'll usually see the signal at month 1-2 of a new cohort — the first-order AOV and 30-day repeat numbers come in below historical norms. That's two to four months before the change shows up in your blended LTV or sitewide revenue report, which is most of the value of running cohorts in the first place.
Track CAC, channels, and funnel conversion in one place
Metricuno connects ad spend, funnel events, and revenue so you can see CAC by channel, cohort, and campaign — without stitching together five tools.