Short answer
Measure repeat purchasing and time to the second order. Build three lifecycle flows around that timing before investing in a points program.
The answer in detail
Start by measuring how many customers return and how long it takes. Then work on the second order with three automated email flows timed to that pattern. Add referrals and a loyalty program after you understand repeat purchasing.
The benchmark
Before choosing a retention tool, establish the store’s observed repeat rate over a defined window. Shopify’s English cohort-retention guide explains grouping customers by shared characteristics such as their first-purchase date and following their behavior over time. It supports the measurement approach, not a universal “normal” repeat-purchase percentage for every category.
Field examples vary widely. A Club winter-accessories brand built around one flagship product had roughly 13% repeat purchasing among customers with a year of observation. An outdoor brand saw returning-customer orders fall during a CRM migration while automated flows were disconnected. That interruption suggests the flows mattered there, but does not quantify their incremental effect.
| Order | Question | Playbook guide |
|---|---|---|
| 1 | How many return, and when? | Repeat purchase rate |
| 2 | When should the second order happen? | Second purchase |
| 3 | What happens immediately after ordering? | Confirmation page and customer account |
| 4 | Which flows and audiences matter? | Automated email flows and attrition |
| 5 | Can customers recruit others? | Referrals |
| 6 | Would points cover their cost? | Loyalty economics |
The rule
Begin with inexpensive work that reaches many customers. Measurement needs an order export, and second-order preparation can begin after each first purchase. Points programs add cost and can reward people who would have returned anyway. A program cannot fix a catalog with little reason to buy again.
Each step supplies the next with information. The observed first-to-second-order interval sets the timing of the replenishment flow. It should not be copied from a tool's default delay or another country's seasonal calendar.
Keep discounts targeted and late in the sequence. Sending every first-time buyer a coupon can subsidize orders that would have happened at full price. This is a hypothesis about incrementality to evaluate, not a claim that discounts never help.
The steps
- Measure a cohort. Select first-time buyers from a defined acquisition period, then divide those placing a second valid order within the chosen window by the full original cohort. Deduplicate identities and exclude marketplace profiles where they do not belong to the web cohort. At one women's-fashion brand, raw email segmentation showed 17% second purchasing while cleaned web-order data showed 25.3%. The definition changed, not the customers.
- Measure timing. Calculate days from first to second order by product family. Use the median alongside the distribution. In the same brand, a 287-day mean was pulled upward by two- or three-year gaps; the median was 139 days. As a starting convention, consider twice the usual repeat interval before classifying a customer as inactive, then check the actual distribution. This practical convention can exceed a year in seasonal categories.
- Fix the immediate experience. Review the confirmation page and account area while attention is high. Explain fulfillment and access to support before introducing another promotion.
- Build three flows. Send product-use guidance after delivery, a repeat-purchase reminder around the relevant interval, and a later message to customers who pass it. Use automation priorities and the attrition definition. Send marketing only to audiences eligible under the applicable market's permission and opt-out rules.
- Introduce referrals. A successful delivery or positive review can be an appropriate moment; a second order is not always necessary. Follow the referral guide.
- Evaluate points last. Consider them when there is already a natural pattern of several annual purchases. Use a holdout to measure incremental contribution rather than comparing members with nonmembers.
The pitfall
A visible points app can feel more concrete than an export. Its most active members may already be the most loyal customers, making a simple member comparison misleading.
Another Club women's-apparel brand had committed €40,000 to a loyalty club that stalled without a clear scope. An audit found 72% of customers had purchased only once, representing 41% of revenue. The immediate priority became the second order, followed by a lighter club built iteratively. Keep the original euro cost as case evidence, not a standard program budget.
What to tell your team
This example is fictional:
“Today, 18% reorder within twelve months, and half of repeat buyers return within fifty-two days. Before a loyalty program, we will build three flows around that interval and compare mature acquisition cohorts, controlling for season and promotion differences.”
