Ecommerce retains about 30% of its customers while banks keep 75% and media companies keep 84%. Those numbers get quoted side by side constantly, and comparing yourself to the wrong column is how stores talk themselves into panic or complacency.
Retention rate is the most quoted and least contextualized number in ecommerce. Somebody reads that the average company keeps 75% of its customers, checks their own dashboard, sees 26%, and concludes their store is dying. It is not. It is a store, and stores live in a different gravitational field than the industries those big retention numbers come from.
Here is the published data, sorted so the comparison is honest.
Shopify’s staff-maintained roundup, updated September 11, 2024, collects average retention rates by industry. The service and contract businesses cluster at the top: media companies and professional services at 84%, automotive and transportation at 83%, insurance at 83%, IT services at 81%, construction and engineering at 80%, financial services and telecom at 78%, healthcare and software at 77%, banking at 75%, consumer services and manufacturing at 67%, hospitality at 55%.
Then, at the bottom of the same table: ecommerce, 30%. That figure comes from Decile’s 2023 Ecommerce Benchmarking Guide, measured across the brands on its platform. Omnisend’s guide cites a nearly identical online-retail average of 28.2%. Two independent datasets, same neighborhood: roughly three in ten customers come back.
Look at what the top of the table sells: subscriptions, policies, contracts, relationships with switching costs. A bank customer who does nothing this year counts as retained. A store customer who does nothing this year counts as churned. Shopify’s own commentary on the table says as much, crediting banking’s 75% partly to “the hassle and costs of switching providers.”
Ecommerce has no such inertia subsidy. Every order is re-earned from scratch against one-click access to every competitor on earth. That is why the honest reading of a 30% store average is not “stores are bad at retention” but “stores measure retention on the hardest setting.” It is also why the moment a store bolts on a contract-like mechanic, a subscription, a loyalty program with meaningful stored value, a replenishment rhythm, its number starts drifting up the table. We cataloged those mechanics in our loyalty program ideas piece.
Click a header to sort. Everything comes from the Shopify roundup except the two rows labeled otherwise.
| Vertical | Avg. retention rate | Source |
|---|---|---|
| Media companies | 84% | Shopify roundup |
| Professional services | 84% | Shopify roundup |
| Automotive & transportation | 83% | Shopify roundup |
| Insurance | 83% | Shopify roundup |
| IT services | 81% | Shopify roundup |
| Construction & engineering | 80% | Shopify roundup |
| Financial services | 78% | Shopify roundup |
| Telecom | 78% | Shopify roundup |
| Health care | 77% | Shopify roundup |
| Software | 77% | Shopify roundup |
| Banking | 75% | Shopify roundup |
| Consumer services | 67% | Shopify roundup |
| Manufacturing | 67% | Shopify roundup |
| Hospitality | 55% | Shopify roundup |
| Ecommerce (all brands, Decile platform) | 30% | Decile via Shopify |
| Online retail | 28.2% | Cited by Omnisend |
A single retention percentage flattens the most important fact about store customers: they are not evenly valuable. Smile.io’s loyalty report, built on a dataset of over 100,000 merchants across 148 countries, estimates that the top 5% of an ecommerce store’s customers generate 35% of its revenue. The Gorgias analysis Omnisend cites makes the same shape with different math: repeat customers are about 21% of a typical store’s customer base but produce 44% of its revenue and 46% of its orders.
So two stores with an identical 30% retention rate can have completely different businesses, depending on whether their retained third buys twice a year or twelve times. The same Smile.io report shows purchase frequency rising across every major ecommerce category in 2024, with CPG brands up 13.95% year over year and recreation up 8.68%, which is a reminder that frequency, not just the retained headcount, is where retention programs earn their keep.
Formula: (end - new) / start × 100. For a store, run it over your natural reorder cycle (90 or 365 days), not a random month. Reference points: 30% (Decile average brand), 28.2% (online retail figure cited by Omnisend).
Industry retention rates are from Shopify's staff-written roundup updated September 11, 2024, which aggregates published figures per industry; its ecommerce line explicitly cites Decile's 2023 Ecommerce Benchmarking Guide (average retention of 30% across brands on the Decile platform). The 28.2% online retail figure and the Gorgias repeat-customer analysis (21% of customers, 44% of revenue, 46% of orders) are as cited in Omnisend's repeat customers guide published January 9, 2024. Revenue concentration (top 5% of customers producing 35% of revenue) and purchase frequency trends (CPG +13.95% YoY, recreation +8.68%) are from Smile.io's State of Ecommerce Customer Loyalty report published February 26, 2025, based on 100,000+ merchants. Caveats worth carrying: each source measures retention over its own window with its own churn definition, the cross-industry figures are directional rather than strictly comparable, and none of these datasets is a random sample of all commerce.
The takeaway is a comparison discipline. Benchmark your store against the ecommerce column, never the cross-industry headline. Beat 30% sustainably and you are ahead of the published field. And when you want to move the number rather than admire it, the levers are the boring ones documented across this site: a welcome flow that converts the first purchase into a second, and a win-back machine that catches the ones drifting away.
Pick a period, then compute (customers at end minus new customers acquired during the period) divided by customers at start, times 100. A store that starts the quarter with 1,000 customers, acquires 400, and ends with 1,250 retained (1250 - 400) / 1000 = 85%... of a customer base that mostly did not need to buy again that quarter. That is why period length matters more in ecommerce than anywhere else: measure over your natural reorder cycle, not a calendar habit.
Against the published baseline, anything sustainably above 30% beats the average brand in Decile's data, and above 28.2% beats the online-retail figure Omnisend cites. But vertical matters: consumables with a monthly reorder rhythm should sit well above the average, while a mattress or luxury store can be perfectly healthy far below it because the honest reorder window is years.
Because most published retention tables mix business models. A bank customer stays by default and leaves by effort; a store customer leaves by default and stays by effort. Subscriptions are the exception that proves it: the moment a store adds a contract-like mechanic, its retention math starts resembling the service industries.
For a store, repeat purchase rate is usually the more honest lens because it does not depend on defining an arbitrary churn boundary for customers who buy on their own schedule. Track retention rate quarterly for trend, and repeat purchase rate by cohort for decisions. We cover the repeat purchase benchmarks separately.