Why some Shopify customers order then return constantly (and how to spot the pattern)

Every DTC merchant eventually notices it. The same customer's name keeps appearing in the returns queue. They placed four orders last month and returned three. They've been a customer for two years and their return rate is 73%. You look closer and find they're not a bot, they're not using stolen cards, and they're not doing anything illegal. They've figured out your return policy and are using it, order after order.

This is the biggest hidden cost in most Shopify stores. A small group of customers (typically 1% to 3% of your total) drives 30% to 50% of your return losses. "Serial returners" is the industry term. Every DTC merchant has them. The question is whether you know who they are.

Why this pattern exists

Return policies evolved to be generous because free, frictionless returns lift conversion. Amazon trained a generation of online shoppers to expect this. DTC brands had to match Amazon's policy to compete for first-time customers.

Most customers use generous return policies sparingly and legitimately. They order once, return if it doesn't fit, move on. A small subset does something different:

• Orders multiple variants of the same item knowing they'll return most (bracketing)
• Orders items for specific occasions and returns after use (wardrobing)
• Accumulates store credit from returns and cycles it into new orders they also return
• Returns items by claiming damage or non-receipt without evidence
• Uses the return window as an interest-free rental

Each behavior is individually explainable. When the same customer shows all of them over 90 days, the pattern is unambiguous.

The 5 patterns you'll see in the data

1. The bracketer

Orders 3 to 5 variants of the same SKU (different sizes or colors) intending to keep one. Data fingerprint:

• Multiple orders, same customer, same product, different variants
• Time between orders under 24 hours (ordered all at once)
• Return rate over 70% on those specific orders
• Usually no issue for low-AOV stores, expensive when AOV is above $100

Bracketing is the hardest to judge. Legitimate customers do it occasionally when sizing is unclear. Abuse is when the same customer does it monthly. See Bracketing: when buying five sizes is normal and when it's abuse for the line.

2. The wardrober

Orders for a specific event (wedding, interview, photo shoot), returns within 5 days of receipt as "unworn." Data fingerprint:

• Orders timed around an obvious event calendar (proms, weddings, holidays)
• Returns initiated 2 to 5 days after delivery
• Return reason "didn't fit" or "changed my mind"
• Often expensive items (formalwear, luxury handbags, jewelry)

Common enough that the NRF has a formal term for it. Wardrobing detection goes into the specific photo, fiber, and fold signals.

3. The frequent flyer

5+ returns in the last 90 days. No single order is suspicious, but the accumulation is. Data fingerprint:

• Return count over 5 in a rolling 90-day window
• Return-to-order ratio over 50%
• Customer has been with you 12+ months (not a bust-out account)
• Orders and returns evenly distributed across the period

Often your "best customers" by top-line revenue, worst customers by contribution margin. Revenue hides the cost of the return cycle.

4. The multi-account cluster

One physical person, multiple email addresses, same shipping address or device fingerprint. Data fingerprint:

• Same shipping address used by 3+ customer accounts
• Same phone number on multiple accounts, or no phone (which is its own signal)
• Similar order patterns across accounts (same SKUs, same time of day)
• Sometimes same IP address if you're tracking that

This pattern often emerges after a customer gets blocked on one account and creates a new one. The fraud rings explained post covers the organized version of this.

5. The refund-method switcher

Accepts store credit on one return, then escalates to cash on the next, or disputes the store credit. Data fingerprint:

• Original refund: store credit
• Follow-up ticket: "I never received the item / it was damaged / I want cash"
• Chargeback often follows 30 to 60 days later

This pattern is specifically post-refund, which is why most fraud tools miss it entirely. See Refund method tracking ROI.

How to spot them in your Shopify data (no tools required)

You don't need fraud detection software to find your top offenders. Open a Google Sheet and run this exercise.

Step 1: export customer data

In Shopify Admin, filter customers by "has returns" and export the list. Include customer ID, email, number of orders, number of returns (use a Shopify app or manual count if you don't track this natively), total spent, total refunded, date of first order, and date of most recent return.

Step 2: calculate 3 metrics per customer

• Return rate: returns / orders, for customers with 3+ orders
• Net contribution: total spent minus total refunded minus (returns × average return cost)
• Abuse ratio: returns in last 90 days / orders in last 90 days, for customers with 3+ orders in that window

Step 3: sort and filter

Sort by 90-day return rate descending. Filter to customers with 3+ orders in the last 90 days. The top 20 to 30 customers on the list are your highest-risk cohort.

Open each customer's order history in Shopify Admin. You'll usually see one or more of the 5 patterns above.

The exercise takes 1 to 2 hours the first time. Merchants are consistently shocked at the result. The top 1% of customers by return count typically accounts for 30% to 50% of total refund dollars.

What to do when you find them

Three options per offender.

Option 1: do nothing

Most merchants default to this. The reasoning: "they're still a customer, they still generate revenue, blocking them would be aggressive."

Wrong frame. The question isn't revenue. It's contribution margin. A customer with a 70% return rate generates negative contribution margin on most DTC categories. You're paying them to keep ordering.

Option 2: apply a stricter policy to them specifically

Tag the customer "elevated risk" and apply different rules. Options:

• Store credit only (no cash refund)
• Restocking fee on returns
• Photo required for damage claims
• No free return shipping (they pay their own)
• Manual review of every return

Shopify customer tags plus Shopify Flow plus your return portal's rules gets you there. Shopify Flow templates for return fraud has pre-built automations.

Option 3: block them

For customers with clear abuse patterns and no sign of changing, blocking is appropriate. You're not the BBB. You don't owe a customer the ability to keep extracting value from your store. When to block a customer covers the playbook.

Why most merchants don't do this

Three common reasons:

"We don't want to accuse good customers of fraud." Nobody's accusing anyone. You're tagging a customer for a different policy tier. They don't know it happened.

"It's not worth our time." Usually wrong. The exercise above pays for itself within weeks because the blocked customers were costing you more than they were paying. Run the math on contribution margin.

"We don't have the data." You do. Shopify has the orders, the returns, and the customer history. You just haven't combined them.

When the spreadsheet stops working

The manual approach works until you hit about 200 returns a month. Above that, you need tooling:

• A scoring system that evaluates every return in real time
• Multi-account linkage detection (same address, different emails)
• Cross-shop intelligence (is this customer flagged on other Shopify stores?)
• Historical context (return history, refund method preferences, chargeback likelihood)

This is what RefundSentry does. Every return is scored against 50+ signals in under 2 seconds. Serial returners are flagged on their first order with you, even if they've never shopped your store before, because the cross-shop identity graph catches them. Free during private beta.

If you're below 200 returns a month, a weekly 30-minute spreadsheet exercise gets you 80% of the value. The important thing is to start. Most merchants never look.

The deeper pattern library: Emerging return fraud patterns. For the most common non-fraud-but-costly pattern, see Gift returns: the fraud vector nobody designs for.