How to Reduce Product Returns
Use a returns survey to find out why products come back, and cut the return rate at its source
Returns are usually treated as a logistics problem. They're a data problem. Every return is a customer telling you that something on the product page didn't match reality — and if you capture what, you can fix the page instead of eating the shipping cost forever.
Most stores already know their return rate. Very few know their return reasons at the SKU level, which is the only version of the number you can actually act on.
Building the survey
Slide 1 — The reason
Add a Single Choice slide: "Why are you returning this item?"
Write options that map to a fix. "Didn't like it" is useless; these are not:
It didn't fit — too small
It didn't fit — too large
It looked different from the photos
The quality wasn't what I expected
It arrived damaged
It arrived too late
I ordered the wrong item
I changed my mind
Splitting "too small" from "too large" matters enormously. A SKU with a 15% return rate that's entirely "too small" is a sizing chart fix. A SKU where returns split evenly between too small and too large is an inconsistent-manufacturing problem. Same rate, completely different response.
Slide 2 — The detail
Add a Long Answer slide: "Tell us more — what would have helped you choose correctly?"
This question is doing the real work. Customers routinely name the exact missing element: a measurement, a photo of the fabric up close, a model with their body type, a note about the fit running small.
Slide 3 — Photo evidence (when relevant)
For quality or damage returns, add a File Upload slide and enable mobile auto-capture so the camera opens automatically on phones.
Use Slide Logic to show this slide only when someone selects "arrived damaged" or "quality wasn't what I expected." Photos of damaged goods let you distinguish a shipping problem from a manufacturing one without a support thread.
Slide 4 — Keep the customer
A return doesn't have to be the end of the relationship. Branch on the reason:
Wrong size → offer an exchange. Use a Redirect To action pointing at the right variant.
Changed my mind → show a Promo Code slide to convert the refund into store credit.
Where to run it
In the returns flow — Trigger the survey when a customer initiates a return with a Custom Survey Trigger. Highest response rate, because they're already in the flow.
Email when a return is received — Send a survey link automatically once the return is processed.
Post-delivery, before a return happens — An order delivery survey asking "How did it fit?" catches sizing problems from customers who are keeping the item. That data is free and it tells you the same thing.
Tag the order and the customer with the return reason using the Add Order Tag and Add Customer Metafield Slide Logic actions on Shopify. The reason then lives in Shopify alongside the order, where your merchandising team already works.
Reading the results
The unit of analysis is the SKU, not the store. A 12% store-wide return rate is meaningless; it's an average of a catalog where a few products are quietly bleeding.
Export responses from the Reports tab, or sync to Google Sheets.
Group by product. Zigpoll captures the Shopify line items associated with each response, so you can see the reason mix per SKU.
Rank products by return rate x margin. Fix the expensive ones first.
Then match the dominant reason to the fix:
Fit (skewed one direction)
Update the size chart; add "runs small" to the product copy
Fit (split both directions)
A manufacturing consistency problem — talk to the supplier
Looked different from photos
Reshoot. Add photos in natural light, on varied body types, with a scale reference
Quality below expectations
Either the copy is overselling or the product is underdelivering. The open-ends will tell you which
Arrived damaged
Packaging or carrier — check whether it clusters by region
Arrived too late
Set delivery expectations honestly at checkout
Z-GPT Insights will summarize the open-ended answers into themes automatically once you have volume, which is how you find the specific missing measurement without reading 400 responses.
Prove it worked
Fix one product page, then watch that SKU's return rate for the next 60 days. Returns are one of the few survey-driven fixes with a clean, direct financial readout — every avoided return is the shipping cost, the restocking labor, and the margin, all recovered.
Related
How to Find Out Why Visitors Aren't BuyingOrder Delivery SurveyFulfillment SurveyLast updated