Reducing your return rate takes more than one number

The return rate is an average, and averages hide patterns. A shop with forty products can have an unremarkable overall rate while two products cause a large share of all returns. Anyone watching only the total will not see those two products. They will only see the number fluctuate.

The rate becomes meaningful once it is available per product and, where variants exist, per variant. A foundation in six shades does not have one return rate but six. If two shades sit well above the others, that is a concrete lead. Perhaps the photos of those shades look warmer than the product appears on skin.

The second gap is the reason. A rate says how often something comes back, not why. Without the reason, every measure remains guesswork. The real work therefore begins with how reasons are recorded.

How to record return reasons so they can be analysed

Most shop systems offer a drop-down list in the returns form. The default entries often read “don’t like it”, “wrong item”, “faulty” and “other”. They are of little use for analysis. “Don’t like it” lumps together exactly what would be interesting: wrong colour, unexpected scent, smaller quantity, different texture.

A useful list is tailored to the range. For skincare products it might consist of five groups. Differs from the description, with sub-points such as colour, scent, texture and quantity. Skin tolerance as a separate point, because it calls for a different response. Transport and packaging for anything that arrived damaged. Ordering error, when customers picked the wrong item themselves. Personal reasons, when the need has simply gone.

An optional free-text field belongs alongside the list. It supplies the wording that is later missing from the product page. If several customers write that the cream is “runnier than expected”, that is almost word for word the sentence the description needs. The form should ask for no more than one required question and one optional field. Every further question lowers the number of answers.

Why customers sometimes give the wrong reason

A recorded answer is not yet an accurate reason. Customers choose the entry that costs them the least effort or gains them the most. That is not bad faith. It is a natural response to the shop’s own rules.

Return shipping costs are the clearest example. If returns are free only for defects, “faulty” is chosen strikingly often. The analysis then shows a quality problem that does not exist. The actual cause, such as a misjudged shade, disappears from the data. “Wrong item delivered” behaves similarly: sometimes it is true, sometimes the customer mixed up the variant.

That is why it pays to compare answers with what arrives in the warehouse. Whoever unpacks the return briefly notes whether the stated reason matches the condition of the goods. Is it really damaged, is it really a different item? This check takes a few seconds per parcel. It turns self-reports into reliable data.

Separating avoidable from unavoidable returns

Not every return is a mistake the shop can fix. Treating them all alike spreads attention across cases where nothing can change. A simple assignment of each reason to one of three groups helps.

Avoidable through information. Better details would have led the customer to a different decision. This covers mismatches in colour, scent, quantity or use. This group belongs on the product page.

Avoidable through processes. Something went wrong between order and delivery. An item was mis-picked, meaning taken from the wrong shelf, or a pump dispenser leaked in transit. This group belongs with the warehouse, the packaging or the carrier.

Unavoidable. The need has gone, a present was bought twice, an individual intolerance could not be foreseen. These returns are part of doing business. They are counted, not fought.

The line is not always sharp. A single intolerance is unavoidable. If reports about one particular product pile up, it becomes a question of information. The assignment is therefore a decision a person makes and reviews regularly.

What returns reveal about the product page

The “avoidable through information” group is the most productive. Every entry in it marks a spot where the page created a false expectation or left a question open. In practice, four patterns keep recurring.

Colours that look different on screen. Product photos are taken under studio lighting and viewed on displays with very different settings. Describing the undertone in words, or comparing it with neighbouring shades, often helps more than another photo.

Quantities that look larger in photos. A 30-millilitre jar can look like a large container when shown on a plain background. A photo in someone’s hand, or a note on how long the quantity lasts with normal use, puts the size in context.

Texture and scent. Neither can be checked online, so both are especially prone to false expectations. Customers’ free-text comments are the best template for wording here.

Details that exist but nobody finds. Sometimes the information is already on the page, just in the third tab below the ingredients. Then the problem is not the text but where it sits. Comparing returns with the questions customer service receives before purchase shows which details belong higher up.

How the findings actually flow back

Many shops record return reasons carefully and still change nothing. The data sits in the system, but nobody is tasked with acting on it. Feeding it back needs a fixed routine, not good intentions.

For a small shop, a monthly half-hour session is usually enough. The person responsible for the range looks at the products with the most avoidable returns. For each one, they decide: adjust the page, change a process or keep watching. Every change is logged with a date. That way, it can later be checked whether it worked.

Only one change per product should run at a time. Replacing photos, text and size details all at once leaves no way of telling what helped. Patience with small numbers matters just as much. For a product with few monthly orders, three returns can be chance. A change is only worthwhile once a pattern holds over several months.

Where product data is maintained centrally, the correction belongs there and not only in the shop. Otherwise the next import overwrites the improved description. When a spreadsheet is enough for this is covered in the article on product data management.

Where the goal of reducing your return rate leads astray

The obvious assumption is that every return avoided is a gain. It is not. A rate can also be lowered by making returns harder: hidden forms, cumbersome procedures, curt replies. The number falls, and so does trust. Customers who expect a return to be a struggle may not order at all when in doubt.

More honest product pages can also mean fewer sales. Stating plainly that a cream is too light for very dry skin will cost some orders. That is intended. Some of those orders would have come back, with shipping, inspection and often unsellable goods as the cost. The yardstick is therefore not the rate alone, but the revenue that stays with the customer.

Finally, the whole routine is not worth it for every shop. With only a few returns a month, an occasional conversation with customer service achieves more than categories and analysis. The system pays off once returns are frequent enough that nobody can keep every case in their head.

After the first round: returns as ongoing feedback

Once the routine is established, the perspective shifts. Returns are no longer just a cost item but regular feedback on the quality of your own product information. New products get a description from the outset that answers the known questions. The list of reasons grows with the range.

The technology behind it is usually simple: an adapted form, a field for the warehouse check, an export for the monthly session. How we build shops with room for routines like this is described under Services.