In a subscription business, churn is easy: the customer cancels. In a Shopify store, customers never tell you they have left. They just stop ordering. So you need a rule that says “after this many days without an order, treat the customer as drifting, at risk, or lost”. That rule is your churn window.
Why “90 days” is usually wrong
Many tools and blog posts use one fixed number, often 90 or 180 days. The problem is that buying cycles differ enormously:
| What you sell | Typical time between orders |
|---|---|
| Coffee, supplements, pet food | Weeks |
| Skincare, haircare, cosmetics | 1 to 3 months |
| Fashion, gifts, home decor | Several months |
| Furniture, electronics, equipment | A year or more |
A 90-day rule would call a perfectly happy skincare customer “lost” too early in some cases, and would never flag a furniture customer who really has gone. The window has to come from your own data.
Measure your median reorder gap
The reorder gap is the number of days between one order and the next order by the same customer. To find your store’s typical gap:
- Take every customer with at least two paid orders.
- Sort each customer’s orders by date.
- Measure the days between each order and the next one.
- Put all those gaps in one list and take the median (the middle value).
Use the median, not the average. A few customers who come back after three years would drag the average far up, while the median shows what is normal.
You need enough data for this to mean something. As a rough guide, wait until you have at least 20 to 30 repeat orders before trusting the number. Until then, start from the table above.
Our free churn window tool calculates the median gap from your Shopify orders export in your browser.
Turn the gap into thresholds
A simple, robust rule:
- Active: last order within 2x the median gap.
- Drifting (medium risk): between 2x and 3x the median gap.
- At risk (high risk): more than 3x the median gap.
Example: if your median gap is 40 days, customers become drifting after 80 days and at risk after 120 days.
Why 2x and 3x? Customers do not reorder like clockwork. Many will be a bit late without being gone. Doubling the median leaves room for normal variation, and by three times the median most customers who were going to come back on their own already have.
Check that the window makes sense
Before you act on it, sanity-check the result:
- How many customers are at risk? If it is 90% of your list, the window may be too short, or you may have a lot of one-time buyers (which is normal for many stores).
- Do known loyal customers look right? Pick five customers you know are regulars and check their status.
- Does it change by product? If you sell both consumables and one-off items, the median may hide two very different groups. Consider separate rules if one product line dominates.
What to do at each stage
- Drifting: a friendly reminder or product tip. No discount yet.
- At risk: a proper win-back sequence, personal for your biggest spenders.
- Long gone: one last-chance message, then stop emailing them.
Let your rule update itself
Your gap changes as your store grows and your product mix changes. WinBack measures your median reorder gap automatically and sets drifting and at-risk thresholds at about 2x and 3x, or lets you pick a preset or your own number of days. It is available on the Shopify App Store with a 7-day free trial.