Ecommerce returns analytics: what to track and how to use it

Most retailers can tell you their storewide return rate. Far fewer can tell you which five products drive most of it, whether "too small" or "too large" dominates a given style, or whether customers who exchange come back more often than customers who refund. That gap is the difference between processing returns and learning from them.

This guide covers what ecommerce returns analytics should track, how to analyse the data, and how to turn it into fewer returns and more exchanges.

What is returns analytics?

Returns analytics is the reporting layer on top of your returns process: the data on what comes back, why, from whom, and what it costs you. At its simplest it's a single return rate. Done properly, it ties every returned item to its product, variant, order, customer, return reason and resolution, so you can ask useful questions instead of staring at one percentage.

Which returns metrics should you track?

A handful of metrics cover most of what you need to know.

Return rate by product, SKU, variant and category, not just storewide
Return reasons with sub-reasons, so "doesn't fit" splits into too small, too large or too long
Resolution mix: how many returns end in a refund, an exchange or store credit
Processing time from lodgement to close
Return value lodged versus closed, so finance can see what's in flight
Repurchase rate after a return, which tells you whether your returns experience keeps customers
Shipping and drop-off performance, so you know where returns get stuck in transit

If your current reporting stops at the first line of that list, you're seeing the cost of returns without any of the causes.

How do you analyse product return rates?

Start at product level rather than store level. A storewide rate of 18% tells you almost nothing, because it averages your best products against your worst. Compare each product's return rate against its category baseline and rank the outliers. Then drill into variants, because a product that looks fine overall can hide one colourway or size with a serious problem.

Sub-reasons turn the outliers into actions. A dress returning at twice the category average with "too small" as the dominant sub-reason needs a fit note on the product page this week, and maybe a pattern change next season. A spike in "faulty" returns that starts right after a restock points at a manufacturing batch, which is a conversation with your supplier rather than your copywriter.

Why context matters in returns data

A return reason on its own is a clue without a crime scene. Knowing that something came back "too small" is only useful when you also know which variant, which order, which channel and where it was returned. Analysts call data like reasons-without-locations a missing context problem, and it's the most common reason returns analysis stalls.

This is mostly determined by how the data is captured. When returns are lodged through a portal that already knows the order, every returned item arrives with its product, customer, reason and location attached, and the analysis becomes straightforward instead of forensic.

How do you use analytics to reduce returns?

Three feedback loops do most of the work. Product page fixes, where sub-reason data turns into fit notes, better measurements and more honest photography. Quality control, where photo evidence attached to faulty returns goes back to your supplier with the claim. And policy, where the data identifies the small group of serial returners so rules can target them without touching anyone else.

We've written a full guide on this in how to reduce customer returns, but the short version is that every reduction strategy starts with knowing which returns are avoidable, and only the data tells you that.

Can analytics increase exchanges?

Yes, and this is the most direct revenue lever in returns data. Your resolution mix shows where customers take refunds when an exchange would have solved their problem. Fit-related returns that end in refunds are the obvious target, because the customer wanted the item, just in a different size.

Once you know where the refunds are leaking, make the exchange path easier than the refund path for those cases. Instant exchanges, where the replacement ships as soon as the return is lodged, shift the mix hard because the customer gets the right item days sooner than a refund-and-reorder would. Then watch the resolution mix in the analytics to confirm the shift is actually happening.

Which returns software provides analytics and reporting?

Look for product-level depth rather than a dashboard with one big number. The test is whether you can answer a specific question, like which variants drive your worst styles, without exporting everything to a spreadsheet first.

Refundid's returns analytics include return rate tables by product, SKU, variant, category and reason, sub-reason breakdowns, resolution mix, processing times, customer return behaviour and repurchase reporting, all in real time. For finance and ops there are CSV exports with more than 50 fields and scheduled reports, and if you want returns data in your own warehouse or BI tool, the API covers it. Our guide to integrating a returns portal covers how that fits into the rest of your stack, and our case studies show what retailers do with it.

FAQ

What is a good return rate for ecommerce?

It depends heavily on category. Apparel and footwear run far higher than most other categories because of fit, so comparing your rate against a generic ecommerce average mostly misleads. The more useful benchmark is your own trend: your rate by category over time, and each product against its category baseline.

How do you analyse return reasons?

Use two levels. A top-level reason like "doesn't fit" plus a sub-reason like "too long" makes the data actionable, and requiring a photo or comment on reasons like "faulty" gives your quality team evidence rather than anecdotes. Then read reasons by product and variant, because a reason that's rare storewide can be dominant on one style.

Can you export returns data into a BI tool or data warehouse?

With Refundid, yes. CSV exports cover more than 50 fields at order or item level, reports can be scheduled, and the API exposes the full returns lifecycle for warehouses and BI tools that want a live feed.

Does returns data help with customer loyalty?

It tells you whether your returns experience is keeping customers or quietly losing them. Repurchase reporting shows how often customers buy again after a return, and how exchange and store credit customers compare against refund customers. We cover that side in how returns drive customer loyalty.

How often should you review returns analytics?

Weekly is enough for most of the year, with two exceptions. Watch daily through January and after any big sale, because peak-season problems compound quickly. And watch new product launches closely for their first few weeks, since early sub-reason data catches a sizing or quality issue before most of the season's stock has sold.