Glossary · cohort analysis ecommerce

What cohort analysis in ecommerce actually tells you about your customers

Cohort analysis in ecommerce groups customers by the date they made their first purchase, then tracks their behavior — repeat orders, returns, unsubscribes — in the weeks and months that follow. It treats each month's new buyers as a separate classroom, measuring which graduating class produces the strongest repeat buyers. A cohort analysis tells you if your customer quality is improving or degrading, not just whether revenue went up.

Think of it like this

An analogy that sticks

Think of a gym owner with a new-member class each January. The January group signs up, comes 8 times their first month, then slowly stops showing up. The March group joins during a different promo, uses the gym differently, and cancels at a different rate. If the owner jams every member into one giant spreadsheet, they can't see the January joiners are 40% likelier to quit by April. They'd keep spending money on promos that bring in fickle members instead of fixing what retains people past Month 3. Cohort analysis is the calendar on the gym wall that separates each starter group so you can see which ones stick.

How it works

The mechanic

You pick a defining event — typically "first order date" — and a metric to track, like repeat purchase rate. Every customer who buys their first item in January goes into the January cohort. In February, you check what percentage of that January cohort bought again. In March, you check again. You do this for 12 months and you have a retention curve that belongs only to that January group.

The January 2024 cohort might have 100 new customers. 18 of them buy again within 30 days. That's an 18% Month 1 repeat rate. Your February 2024 cohort has 120 new customers, but only 12 buy again within their first 30 days — a 10% repeat rate. Revenue is up month-over-month, but you just acquired weaker customers. Maybe your February ad creative brought in discount-chasing one-and-done shoppers.

You'd never see that by staring at a dashboard showing total revenue. The cohort table makes the deterioration obvious: Month 1 repeat rate dropped from 18% to 10% while ad spend rose 15%. Now you have a specific problem to fix — the creative, the offer, or the initial post-purchase email sequence.

The columns of a standard Shopify cohort table run Month 0 (the acquisition month) out to Month 12. Values are usually a percentage of the cohort size, not raw counts, so you can compare cohorts of different sizes. A 100-person cohort with 30 Month 1 repeaters outperforms a 1,000-person cohort with 200 Month 1 repeaters.

Why brand owners care

The business outcome

A $4M skincare brand running Shopify and Klaviyo can have a 35% blended repeat purchase rate and think retention is humming. Cohort analysis often tells a harsher story: their repeat rate was 44% for cohorts from Q1 2023 but has declined to 28% for Q4 2023 cohorts. The average looks fine because older, good cohorts pull it up. Meanwhile, the newest cohorts are getting worse and nobody is adjusting the welcome flow, the post-purchase email, or the retargeting audiences.

This matters straight to gross margin. Acquiring a new customer can cost $40 on Meta. Getting an existing customer to buy again often costs $5 in Klaviyo sends. If your newest cohorts are shedding repurchasers, your overall blended acquisition cost rises because you're replacing good customers with more expensive new ones.

Brands that fix a degrading cohort within 60 days usually see a 15-20% improvement in that cohort's lifetime value (Mailchimp benchmark, 2024). The fix is rarely one email — it's adjusting the offer, changing the onboarding sequence, or cutting an ad set that's bringing in the wrong buyers. Without a cohort view, none of those decisions get made.

In your stack

How to actually do it

Shopify Analytics has a native cohort report under Analytics > Reports > Customer cohort analysis. It shows you the percentage of customers who returned, by the month they first bought. This is the fastest way to check for a deteriorating pattern. The limitation is it doesn't break out by channel or product — it's all first orders aggregated.

For Klaviyo users, build a segment of "customers who had their first order in January 2024" using the "First Placed Order" property. Then pull that segment into a repeat purchase metric over the following months via Shopify's order export or Klaviyo's own analytics if you're tracking second-purchase events. The Klaviyo path lets you split cohorts by acquisition source — did your TikTok organic buyers repeat better than your Meta ad buyers? That one insight can shift $50,000 in ad spend.

If you're doing this in Google Sheets, export your Shopify order sheet, pivot by customer ID, find each customer's earliest order date, and group them into monthly buckets. Then count how many made a second order within set intervals — 30, 60, 90 days — and divide by cohort size. You don't need SQL to get the high-signal stuff.

A worked example

Applied to a real store

A $6M DTC home goods brand using Meta, Klaviyo, and Shopify ran a cohort analysis they'd been putting off. Their CAC on Meta had crept up from $34 to $41 over four months. Blended AOV was flat at $85. The team assumed rising ad costs were just competition and kept feeding the same prospecting campaigns.

The cohort table showed their July 2024 cohort of 1,200 new customers had a 22% Month 1 repeat rate. The October 2024 cohort of 1,100 customers dropped to 11%. Same ad creatives, same landing pages, same welcome flow — but the customers were fundamentally different. The October group had an average first-order discount rate of 28% vs. 14% in July. The Meta campaign was optimizing toward cheaper clicks, which meant coupon-hunters who never came back.

They duplicated the campaign, changed the optimization goal to value-based bidding with a 1% Lookalike from purchasers, and cut the introductory discount from 20% to 10%. The December cohort's Month 1 repeat rate rebounded to 19%. CAC settled at $35. The fix was $200 in ad spend tests and an hour reconfiguring Klaviyo's post-purchase flow. Without the cohort view, they'd still be blaming Meta's auction and raising prices.

The retention curves between the July and October cohorts diverged at Month 1 and never re-converged. That gap over 12 months projected to a $280,000 difference in lifetime value from two adjacent quarterly cohorts. The problem was visible in the back half of the cohort table within 45 days.

Watch out

Common mistakes

  • Measuring blend instead of cohort. A 30% average repeat purchase rate often hides a 15% rate on the newest cohort and a 45% rate on the oldest. You fix what you can see, and you can't see the newest problem in the average.
  • Using too long an interval. A quarterly cohort instead of monthly means you won't spot a deteriorating pattern until 90 days of bad customers are already in your email list. Monthly is the right granularity for a $3-10M brand.
  • Tracking opens instead of purchases. Email engagement is an input, not the outcome. A "Post-Purchase" Klaviyo flow with 45% open rates is meaningless if that cohort's actual second-order rate is 8%. Always tie cohorts back to Shopify order data.
  • Ignoring zero-party attribution. If 40% of a cohort's second purchases come through a channel you're not tagging — like direct site traffic or SMS — the cohort table will mislead you by underreporting repeat rates. Tag anywhere you can and reconcile Shopify's order source data.
See also

Related terms

  • rfm-analysis
  • customer-lifetime-value
  • churn-rate-ecommerce
  • klaviyo-segmentation
  • customer-acquisition-cost
Plain English

cohort analysis ecommerce in two sentences

Cohort analysis in ecommerce groups customers by the date they made their first purchase, then tracks their behavior — repeat orders, returns, unsubscribes — in the weeks and months that follow. It treats each month's new buyers as a separate classroom, measuring which graduating class produces the strongest repeat buyers. A cohort analysis tells you if your customer quality is improving or degrading, not just whether revenue went up.

A $4M skincare brand running Shopify and Klaviyo can have a 35% blended repeat purchase rate and think retention is humming. Cohort analysis often tells a harsher story: their repeat rate was 44% for cohorts from Q1 2023 but has declined to 28% for Q4 2023 cohorts. The average looks fine because older, good cohorts pull it up. Meanwhile, the newest cohorts are getting worse and nobody is adjusting the welcome flow, the post-purchase email, or the retargeting audiences. This matters straight to gross margin. Acquiring a new customer can cost $40 on Meta. Getting an existing customer to buy again often costs $5 in Klaviyo sends. If your newest cohorts are shedding repurchasers, your overall blended acquisition cost rises because you're replacing good customers with more expensive new ones. Brands that fix a degrading cohort within 60 days usually see a 15-20% improvement in that cohort's lifetime value (Mailchimp benchmark, 2024). The fix is rarely one email — it's adjusting the offer, changing the onboarding sequence, or cutting an ad set that's bringing in the wrong buyers. Without a cohort view, none of those decisions get made.

FAQ

Common questions

  • What is cohort analysis in ecommerce?

    Cohort analysis in ecommerce groups customers by the time they first purchased, then tracks how many buy again in each following month. It answers: are our new customers getting better or worse month over month? A brand might find their January cohort has a 22% Month 2 repeat rate while the March cohort only hits 14% — that gap is a real signal about inventory, ads, or onboarding.

  • What is the 80/20 rule in ecommerce?

    The 80/20 rule in ecommerce means roughly 80% of revenue comes from about 20% of customers — your repeat buyers. Cohort analysis proves or disproves this for your particular brand. Some $5M brands discover their top 8% of customers (often a "Premium Repeat" archetype) generate 60% of gross profit. That group should define your retention campaigns, not your new-customer discount ads.

  • What is customer cohort analysis in Shopify?

    Customer cohort analysis in Shopify uses the platform's built-in reports or connects your order data to tools that group buyers by their first order month. Shopify's native cohort report in the Analytics dashboard shows you the percentage of customers who return to purchase again over time, broken out by acquisition month. It's a quick way to see whether your repeat buyer program is gaining or losing steam.

  • How do I run a cohort analysis for a Klaviyo-based Shopify store?

    You need two pieces: the date of each customer's first order, and their order dates after. Shopify's API gives you this, and Klaviyo can segment by "First Placed Order" date. Build a month-one cohort segment in Klaviyo and check how many recipients placed a second order within 30, 60, 90 days using Shopify's order export matched back to the segment. Don't just measure open rates on the flow — measure the downstream purchase rate.

Run the audit

See cohort analysis ecommerce on your own store.

Free. Seven-minute connect. About 24 hours to your six named buyer archetypes plus 18 ranked campaigns.