Updated: Nov 3
How to Measure Customer Retention in Ecommerce: Beyond Repeat Purchase Rate

Repeat purchase rate is useful, but it is not a complete retention strategy. Retention is the set of behaviours that make a valuable long-term customer more likely to exist. The second purchase is one important milestone, not the whole system.
Start before the repeat purchase
A customer does not become valuable at the moment the dashboard records order number two. The journey starts with who was acquired, what promise brought them in, what they bought first and whether they actually experienced the product's value.
For many categories, the strongest early retention signal is behavioural: activation, usage, product adoption, a completed setup, a streak, a replenishment cue or another action that happens before the next transaction.
The six layers of retention measurement
1. Customer quality
Which acquisition sources, promises and entry products create customers with the strongest downstream economics?
2. First-purchase economics
What margin did the first order create, and what must happen next for the customer to become profitable?
3. Activation and adoption
Did the customer actually use the product, receive value and form the behaviour that precedes repeat purchasing?
4. First-to-second purchase
How many customers buy again, how quickly, what do they buy, and at what margin?
5. Purchase path
Which second and third purchases are associated with the strongest long-term value?
6. Lifespan and contribution
How long does the relationship persist and how much contribution margin does it create over time?
Why first-to-second purchase deserves special attention
The first purchase proves that acquisition worked. The second purchase begins to prove that the customer relationship has value beyond the initial conversion.
But the quality of the second purchase matters. If a customer simply repeats an unprofitable entry product, repeat purchase rate can improve while the economics remain poor.
Study your best customers, not only the average
Average reorder time tells you what normally happens. Your best customers can tell you what is possible.
Compare the top customer cohorts with the average. What did they buy first? How quickly did they return? What did they buy second? Which behaviours happened between orders?
The goal of lifecycle marketing is not merely to predict the average. It is to identify valuable behaviours and make them more likely.
Separate weak retention from a missed opportunity to buy
A customer who cannot reorder an unavailable product still has no repeat purchase in the reporting window. That is a real business outcome. It is not, by itself, evidence that the customer disliked the product or that the email programme needs a stronger offer. Availability changes how the result should be interpreted, not whether the missed order occurred.
Keep the observed cohort result intact
Fix the qualifying customer population, order definition and observation window before reviewing performance. Show the actual first-to-second-purchase rate for that cohort. Do not quietly remove customers affected by stockouts, pause their calendar age or move late orders backwards to make the 90-day result look healthier.
In a fictional cohort of 100 first-time customers, 20 place a qualifying second order within 90 days. The observed 90-day rate is 20%. If another 30 return after day 90 when stock becomes available, the later cumulative rate can reach 50% once every customer has the same longer observation window. The original 90-day rate remains 20%. Neither figure isolates the effect of an email.
Add an availability record beside the outcome
For the relevant first product or repeat-purchase category, record the market, eligible variants, historical stock periods, fulfilment restrictions, promotion dates and acquisition changes. Use inventory history where available. Today's catalogue cannot establish whether a product was buyable throughout a past cohort's window. Mark missing history as unknown.
An availability-based breakdown can help investigate the cause. Keep it secondary to the original result, state its inclusion rule and show the customer counts. Customers buying different products or seasons can differ in many other ways, so a higher repeat rate among those with stock available is not automatically the causal effect of stock availability.
Distinguish the recovery from the baseline
After stock returns, report the affected older cohort separately from newly acquired customers. A burst of restock orders may represent delayed demand rather than a repeatable improvement in customer loyalty. It can also change the apparent share of campaign and flow revenue without showing that one part of the programme suddenly became more effective.
Use this diagnosis to choose the next action: repair supply or delivery constraints, improve product adoption, change the reorder timing, or test a specific message. The replenishment guide covers the interrupted customer journey. The attribution guide explains why reported channel credit does not settle the causal question.
Metrics worth tracking
First-to-second purchase rate
Time to second purchase
Second-purchase product mix
Cohort contribution margin
Customer lifespan
Purchase frequency
Activation or usage milestones
Repeat rate by acquisition source
LTV by entry product
Reactivation rate for lapsed customers
The principle
Purchase frequency is an outcome. Retention is the system of behaviours, experiences and interventions that make that outcome more likely.
If you only measure transactions, you often notice the problem after the customer relationship has already weakened.
About Thrivelia
Thrivelia is a DTC retention and LTV growth agency helping ecommerce brands improve lifecycle strategy, CRM, email, SMS, customer economics and long-term value.
Continue with retention marketing services and Thrivelia client work.
Connect retention measurement to the next decision
Use the post-purchase email strategy guide to turn the cohort finding into a useful second-purchase journey.
Our email marketing attribution guide distinguishes reported credit from additional customer purchasing.
The programmatic direct mail example shows how a positive revenue result can still fail to cover the cost of an extra channel.

.png)
Comments