Churn Diagnosis
Customer retention diagnosis
Diagnose a retention problem by defining eligible repeat purchases, checking buying cycles and investigating plausible causes before choosing an action.
Customer retention diagnosis starts with one precise question: among customers who had a reasonable opportunity to buy again, what changed in their purchasing, and what might explain it? A quiet order history is an observation only. It does not prove a customer has left or that a campaign will bring them back.
Define the purchase you expect to repeat
Choose the customer group, product category and follow-up period before calculating a rate. Count completed purchases consistently, and decide how to treat cancellations, refunds, exchanges and subscription renewals. If a buyer moves between pack sizes or close substitutes, an identical-product count may miss a genuine repeat purchase.
Start with a meaningful first purchase and give customers the same amount of time to return. A buyer acquired last week cannot fairly be compared with one acquired last year on whether they made a second purchase within six months.
Check whether customer records join online and in-store orders reliably. A missing link can make a returning buyer look new.
Buying opportunities differ by category. A household consumable may be needed again soon; a durable item may not.
Separate products with different buying rhythms before treating a longer gap as a retention problem. Keep customers whose next purchase opportunity is unclear in an unknown group.
Read several signals together
No single number identifies the cause of fewer repeat sales.
| Question | Useful observation | What it cannot establish alone |
|---|---|---|
| Are first-time buyers returning? | Share of a mature first-purchase group making a qualifying second purchase within a fixed period | Why the others have not bought |
| Are established buyers slowing? | Time since the last relevant order compared with suitable past buying gaps | Whether an individual has permanently left |
| Is the commercial result changing? | Completed repeat orders and net sales, with contribution assessed only where costs are reliable | Whether a retention action caused the change |
| Is there a plausible experience problem? | Recurring delivery, product or support issues in order and case records | That every affected customer stopped buying for that reason |
Check the denominator as well. A larger share of orders from returning customers can reflect fewer new customers rather than stronger repeat behaviour. Report returning and eligible customer counts alongside any percentage.
Key Metrics in Customer Retention Diagnosis
- Share of returning buyers
- Percentage of first-time customers making a second purchase within a fixed period
- Time since last order
- Compared against historical buying gaps to detect slowing behaviour
- Net sales from repeat orders
- Revenue after discounts and reversals; used to assess commercial impact
- Eligible customer count
- Number of customers with opportunity to return; essential for denominator accuracy
- Return rate trend
- Change over time in repeat purchase rates; indicates potential systemic issues
Investigate the pattern before choosing an action
Compare mature groups whose first purchases occurred in comparable periods. Use the same elapsed follow-up time and qualifying-purchase rule.
If a decline appears in one category, region or first-order route, investigate that group. Compare like seasons where buying opportunities vary through the year.
Review a sample of order histories for bulk purchases, returns, substitutions, stock gaps and subscription deliveries that may explain longer intervals. Read the underlying service records before assigning a cause from case tags.
Some unhappy customers never contact support, while frequent buyers may have more occasions to do so.
Write down competing explanations. A fall in second purchases after a product launch might reflect a product issue, a longer use cycle, missing linked orders or a change in the mix of first-time buyers.
State what evidence would strengthen or weaken each explanation. An association between complaints and inactivity warrants investigation; it does not establish that the complaints caused every later gap.
Pros and Cons of Using Case Tags to Identify Retention Issues
- ProsCan reveal recurring delivery, product or support issues; useful for spotting patterns
- ConsDoes not prove all inactive customers were affected; some unhappy customers never contact support
Check what the report actually counts
Before treating a dashboard as current, check its data freshness. Shopify's Customer reports might not display activity from the past 12 hours, while its New vs returning customer report is up to date, give or take a few seconds.
Refreshing a report can display newer data.
Check whether the reporting period refers to the date of an order or the period in which a customer is classified. Shopify's customer reports use a customer's entire order history, not only orders placed during the selected timeframe.
For example, a customer whose first order was in November can appear as returning in a November report if their second order is placed in December.
Shopify defines a first-time customer as someone who placed their first order with the store, and a returning customer as someone whose order history already includes at least one order.
Its New vs returning customers report displays the number of customers in each group, so distinguish that count from the number of orders they placed.
Use a consistent sales measure when checking whether repeat purchasing changed. Shopify defines net sales as gross sales minus discounts and sales reversals, and describes it as an approximation of actual revenue.
Total sales includes additional amounts such as fees, duties, shipping charges and taxes, so it answers a different question from net sales.
Shopify Customer Report Definitions: New vs Returning Customers
- New CustomerSomeone who placed their first order with the store
- Returning CustomerSomeone whose order history already includes at least one order
- Reporting Period FocusCustomer's entire order history, not just orders within the selected timeframe
- Net Sales DefinitionGross sales minus discounts and sales reversals (approximation of actual revenue)
- Total Sales InclusionIncludes fees, duties, shipping charges, and taxes – answers a different question than net sales
Match the response to the finding
If evidence points to a recurring operational failure, assign an owner and check whether the repair reduces it. If a gap falls within the product's usual buying window, continue ordinary service rather than labelling the customer lost.
If records cannot reliably identify repeat orders, repair the measurement before setting a campaign target.
For a proposed reminder, service change or offer, define the eligible group, intended outcome and review period in advance. Later purchases show what happened.
Estimating whether the action added purchases requires a suitable comparison group that did not receive it.
Keep a short record of the customer and product group, purchase rule, follow-up window, observed change, alternative explanations, evidence reviewed and next check.
Revisit it when the product mix, data or buying cycle changes.
In this guide
- Retention versus loyalty: defining the business problemLearn what repeat purchases can show, what loyalty requires beyond order data, and how to write a useful customer problem statement.
- Choosing a repeat-purchase measure for a product categoryChoose a category-appropriate repeat-purchase metric with a clear numerator, eligible customer group, purchase definition and follow-up window.
- Separating customer loss from naturally infrequent buyingAssess a quiet customer history against product buying opportunities, comparable purchase gaps and complete observation windows.
