Most advice on how to increase repeat purchases in ecommerce starts with email and loyalty points. Those can help. But a customer's decision to buy again is mostly made before the second email arrives: in the product, the delivery, the first use, and whether your range gives them a reason to come back.
So the first question is not which retention flow to build. It is how often a satisfied customer would naturally buy what you sell, and what gets in the way.
Understand the natural repurchase cycle
Different products run on different clocks. Treating them the same is how stores end up sending replenishment reminders for a sofa.
| Product type | Examples | What a repeat looks like | Main lever |
|---|---|---|---|
| Consumables | Coffee, candles, pet food, skincare | The same or a similar product, when it runs out | Timing and convenience: reminders, subscriptions, easy reorder |
| Durable goods | Furniture, eyewear, kitchenware | Rarely the same item. A second product, a gift, an upgrade years later, or a referral | Range, experience and referral, over a long cycle |
| Accessories and add-ons | Filters, cases, refills, spare parts, care products | A companion to something the customer already owns | Relevance to the original purchase and the timing of need |
Consumables. Estimate the usage interval from the product itself: how many days a bag, bottle or candle lasts at normal use. Then check it against your order data, looking at the typical gap between first and second orders among customers who did reorder.
Durable goods. A low repeat rate within a year can be normal and healthy. The useful questions are whether you have a second product that makes sense, whether customers recommend you, and whether the business model quietly assumes repeat purchase the product cannot produce.
Accessories. The opportunity is usually tied to a moment, such as a filter due for replacement or a change of season. Relevance matters more than frequency. An offer for the wrong model of spare part is worse than no email at all.
Before you set a repeat-purchase target, write down how often a happy customer would need what you sell. If the answer is once every three years, a 90-day repeat rate is the wrong measure.
Improve the first purchase and follow-through
The second order is mostly earned in the first. A customer who waited too long, received a damaged item or could not work out how to use the product is unlikely to return, however good the follow-up emails are. Many of the biggest causes of low repeat purchase have nothing to do with email.
-
Product quality and fit.
If returns or complaints cluster around one product, it may be costing you more repeat customers than any email can win back.
-
Delivery experience.
Late, unclear or damaged deliveries sour the first impression. Check actual delivery times against what the site promised.
-
Stock-outs.
A customer who comes back and finds their product unavailable may buy a substitute from someone else, and stay there.
-
Unclear use or results.
Skincare that takes weeks to show results, or equipment with a learning curve, needs guidance at the right moment, or customers decide it did not work.
-
A thin range.
If there is nothing sensible to buy next, retention marketing has little to sell.
-
The wrong first customer.
Heavy first-order discounts can attract buyers who came for the deal. They may not return at full price, and that shows up months later in the numbers.
-
Service when something goes wrong.
How a problem is handled often decides whether a customer comes back.
Follow-up messages help with the follow-through: clear order and delivery updates, how-to-use content timed to when the product arrives, and a check-in once the customer has had time to use it. Which email flows to build first covers how to prioritise those.
Repeat purchase is earned in the first order, not requested in the fifth email.
Compare replenishment, cross-sell and loyalty ideas
Once the first experience is sound, choose tactics that fit the product type. Each one has a cost as well as a benefit.
| Tactic | Suits | What it costs or risks |
|---|---|---|
| Replenishment reminders | Consumables with a predictable usage interval | Irritating if the timing is wrong; the interval should come from real reorder data |
| Subscriptions | Consumables used at a steady rate | Discount given on orders that may have happened anyway; cancellation friction damages trust |
| Cross-sell and bundles | Stores with complementary products | Poor matching feels like spam; bundles can lower margin if priced carelessly |
| Accessory follow-ups | Durable products with add-ons or parts | Needs accurate data on what the customer owns |
| Loyalty points or tiers | Frequent-purchase categories with healthy margins | Points are a delayed discount, and often reward customers who would have returned anyway |
| Referral programmes | Considered or durable products people talk about | Reward cost and some abuse risk; only works if the product is worth recommending |
| Range extension | Stores where customers keep asking what else you sell | Stock, cash and development cost; the biggest commitment on this list |
A useful test for any tactic: which customers change their behaviour because of it? A loyalty programme that mostly rewards existing regulars can raise the cost of orders you were already getting without adding many new ones. Improving profit without chasing more orders looks at retention through that contribution lens.
Matching the tactic to the product
A store sells a $240 water filter jug, which is a durable purchase, and $35 replacement cartridges that last about two months, an accessory with a consumable cycle.
A generic “shop again” email to the jug buyer a month later has little purpose. A reminder about seven weeks after delivery, naming the cartridge that fits their jug, matches the natural cycle. A cartridge subscription may suit some customers, but only if cancelling is easy.
The retention opportunity sits in the cartridges, not the jug. The plan should say so, and the measure should follow it.
Measure cohorts rather than relying on a headline
A single repeat purchase rate for the whole store mixes customers acquired years ago with customers acquired last week. It can rise or fall for reasons that have nothing to do with anything you changed.
A cohort groups customers by when they first bought, then tracks what share of each group buys again within a fixed window, such as 90 or 180 days. Every group gets the same amount of time, so you can compare them fairly.
| First-purchase month | New customers | Ordered again within 90 days | Repeat rate |
|---|---|---|---|
| Month 1 | 200 | 36 | 18% |
| Month 2 | 180 | 27 | 15% |
| Month 3 (discount campaign) | 320 | 32 | 10% |
| Month 4 | 220 | 44 | 20% |
| All four months | 920 | 139 | 15.1% |
Hypothetical figures for illustration. Only cohorts whose full 90-day window has passed are included.
The combined 15.1% hides the story. The discount month brought in the most customers and the lowest repeat rate. Month 4 did better, which is worth investigating, but one month is not proof that any particular change caused it.
Once your volume allows, cut cohorts by first product, acquisition channel, whether a discount was used and region. Track the time between first and second orders, not only whether a second order happened. At low volumes, a few customers either way can move a cohort's rate a long way, so look for patterns across several months.
Do not let assumed lifetime value fund today's spend
Lifetime value is a forecast. It is tempting to use it to justify paying more for a new customer than the first order returns. That can work for a business with proven repeat behaviour and the cash to wait. For most stores, it hides a loss that has to be financed now.
The gap an optimistic lifetime value hides
A store's first order returns $35 of contribution after product, shipping and fees. It pays $60 in media to win each new customer, reasoning that customers buy three times, so each is worth $105.
Its cohort data shows 20% of customers buy a second time within six months. Six months in, the average customer has returned $35 + (20% × $35) = $42 of contribution against $60 of acquisition cost: still $18 short.
Winning 500 customers this way costs $30,000 of media up front and returns $17,500 of first-order contribution. The $12,500 gap has to come from cash the business already has, while it waits to find out whether the third purchase ever happens.
Set acquisition limits from what the first order, or a proven early repeat pattern, returns. Treat later purchases as upside you measure, not a budget you have already spent. Why ecommerce sales plateau covers what happens when a store has been leaning on that upside for too long.
If retention turns out to be your lever, our email and retention work explains how we build it into a wider plan. More guides on growth and retention are in the resources hub.
You are on the list. The next one lands on Friday.
About the author
Mareike NiedermeierMareike Niedermeier is the founder and CEO of Sales Savvy Online and the creator of Survivable Distance, a theory of ecommerce viability. Since 2019 her team has worked with more than 200 ecommerce brands. The Australian Business Journal named her a Top Ten Marketer in Australia.
More about Mareike →