How to Calculate Ecommerce Customer Retention Rate: Formula and Examples

Calculate ecommerce customer retention rate with this formula: ((customers at the end of the period minus new customers acquired during the period) divided by customers at the start of the period) x 100. A useful ecommerce customer retention report must also define what counts as an active customer and choose a period that matches the product’s normal repurchase cycle.

For example, if a store starts with 2,000 eligible customers, has 2,300 active customers at the end, and 800 of those are new, retention is ((2,300 – 800) / 2,000) x 100 = 75%. The arithmetic is simple. The difficult part is building comparable customer sets.

Quick answer: what is the ecommerce retention formula?

Customer retention rate (CRR) = ((E – N) / S) x 100

• E: customers at the end of the measurement period
• N: new customers acquired during that period
• S: customers at the start of the period

Shopify gives the same formula in its guide to customer retention programs and metrics. Keep one definition for E, N and S in your reporting documentation so monthly results remain comparable.

How do you calculate retention step by step?

1. Choose the measurement period, such as a month, quarter or year.
2. Define an active customer for your business.
3. Count eligible customers at the start of the period.
4. Count active customers at the end of the period.
5. Identify which end-period customers were acquired during the period.
6. Subtract new customers from end-period customers.
7. Divide by start-period customers and multiply by 100.

Worked example for a non-subscription store

Suppose a personal-care store measures quarterly retention. It defines an active customer as someone who purchased during the relevant 90-day window.

First remove new customers from the ending count: 2,300 – 800 = 1,500 retained customers. Then divide by the starting count: 1,500 / 2,000 = 0.75. Multiply by 100: customer retention rate = 75%.

This means 75% of the defined starting customer group met the activity condition in the ending period. It does not mean every one of those customers bought on the final day, stayed subscribed, or ordered the same product.

Why is ecommerce retention harder to define than subscription retention?

A subscription business has a visible active or canceled state. Most retail customers do not announce that they have left. A person who has not bought coffee for 60 days may be inactive, while a laptop customer who has not bought for a year may be entirely normal.

LoyaltyLion’s ecommerce retention guide makes the same practical point: reporting periods must account for purchase frequency and seasonality. It notes that a yearly comparison can be too short for laptops and too broad for books. See its discussion of ecommerce retention formulas and benchmarks.

For non-contractual ecommerce, document one of these activity rules:

• Placed at least one order in the current period
• Placed an order within the last X days
• Placed another order within X days of the first purchase
• Belongs to a first-order cohort and repurchased by a defined age

The third and fourth options often produce more actionable retention analysis because every customer receives a comparable opportunity to return.

How do you choose the reporting window?

Use the distribution of days between first and second orders. Find the median and the point by which a meaningful majority of second purchases normally occur. Build reporting windows around those patterns and the actual use cycle of the product.

These are analysis starting points, not industry percentage benchmarks. Validate them against the store’s own time-to-second-order curve.

What is a good ecommerce customer retention rate?

A good rate is one that improves against a comparable cohort, period and product mix while maintaining contribution margin. Universal percentages can mislead because the definition and repurchase opportunity vary widely.

Use a benchmark ladder:

1. Internal trend: compare with the previous equivalent period using the same definition.
2. Cohort trend: compare customers acquired from similar channels, offers and first products.
3. Category benchmark: compare products with similar natural replenishment cycles.
4. External benchmark: use only when the source defines the formula, period, customer type and dataset.

Shopify has cited a 30% average across ecommerce in a broader metrics article, but that figure should not become an automatic target for every store. A rate without a stated window and customer definition is not directly comparable. The better management question is: “Are like-for-like cohorts returning more often and more profitably?”

Customer retention rate versus repeat purchase rate

These measures are related but answer different questions.

Do not label returning visitors as retained customers. A visitor can return without ordering, and a buyer can return on another device. Customer metrics should use a stable customer or order identity where permitted.

How does Shopify classify new and returning customers?

Shopify’s customer reports documentation defines a first-time customer as someone placing their first order and a returning customer as someone whose order history already includes an order. It also warns that some customer report data can reflect a customer’s entire later order history, not only activity visible at the selected time.

This matters when exporting data. Confirm whether the platform classifies customers using information known at the time of the report or their current lifetime order history. Save the query logic and extraction date with every dashboard.

How do you build a cohort retention table?

Group each customer by first-order month. Then count the percentage of that original group that purchases again at month 1, month 2, month 3 and so on.

These figures are illustrative. The table shows an improvement at equal cohort ages, which is more informative than comparing March’s mature customer base with a January cohort that had longer to repurchase.

Which data problems distort retention rate?

• Guest checkout duplicates: one buyer appears under multiple email addresses or customer IDs.
• Refunded and test orders: non-revenue activity remains in the customer count.
• Wholesale and retail mixing: very different purchase cycles share one denominator.
• Subscription and one-time orders mixing: contractual and discretionary retention are combined.
• Partial period comparison: an incomplete month is compared with a complete month.
• Changing active-customer definition: dashboards move without real behavior changing.
• Seasonality: gift-acquired customers are compared with everyday cohorts.
• Insufficient cohort age: recent buyers have not had a fair chance to reorder.

Clean the identity and order rules before trying to explain small movements. A precise formula applied to inconsistent data still produces a misleading result.

What should you segment after calculating CRR?

Break retention down by first product, first-order discount, acquisition source, country, fulfillment outcome and first-order value. Look for material groups, not every possible slice. A channel can acquire many customers at a low first-order cost while producing weak 90-day retention. A product can convert well but create returns or poor second-purchase behavior.

Link the result to ecommerce email marketing and service data. If a cohort has lower retention, examine delivery exceptions, product returns, support reasons, message eligibility and days to first follow-up before assuming the problem is the offer.

How often should retention rate be reported?

Update an operating dashboard monthly, but use rolling and cohort views that match the repurchase cycle. A consumables store might report 30-, 60- and 90-day second purchase monthly. A durable-goods store might emphasize quarterly accessory purchase and annual returning-customer behavior.

Do not declare a strategy successful before the cohort reaches the selected age. Label incomplete cohorts clearly and exclude them from like-for-like summaries.

A retention measurement checklist

1. Write the business definition of an active customer.
2. Choose a period based on actual purchase intervals.
3. Exclude test, canceled and fully refunded orders consistently.
4. Deduplicate customer identities according to privacy and data rules.
5. Calculate CRR with documented E, N and S counts.
6. Add repeat purchase and cohort retention views.
7. Segment by first product and acquisition source.
8. Compare contribution margin, not only customer counts.
9. Keep the method unchanged when tracking trends.
10. Annotate promotions, stockouts and tracking changes.

FAQs

Can ecommerce retention rate be over 100%?

The standard customer-count formula should not exceed 100% when E, N and S are defined as comparable sets. A result above 100% usually indicates inconsistent periods, duplicate identities or an incorrect new-customer count. Revenue retention can exceed 100%, but that is a different metric.

Is retention rate the opposite of churn?

They can be complements in a subscription model with clear active and canceled states. In non-subscription retail, inactivity is harder to define, so do not assume customer churn equals 100% minus CRR without matching definitions.

Should refunded customers count in retention?

Set a consistent rule. Many stores exclude fully refunded, canceled, fraudulent and test orders because they do not represent completed customer value. Document the rule and apply it to every period.

What period is best for ecommerce retention?

Use a period that gives customers a realistic chance to repurchase. Review days between first and second orders, seasonality and product life. Report multiple cohort ages when one window cannot represent the catalog.

Which metric is best for non-subscription ecommerce?

Cohort repeat purchase at fixed ages is often easier to interpret than a blended CRR. Use it alongside days to second order, customer lifetime value and contribution margin.