In short. Returns come down when you tackle their cause before the purchase: size, how faithful the product page is, reviews, packaging, order picking, delivery times and promotions that push people to over-buy. All seven start from a survey on why people return; three also rest on studies that measure their effect. None of them adds friction for the customer who has already decided to return, which in the European Union is the red line.
Almost every list on how to reduce ecommerce returns shares two flaws. The first: they mix proven tactics with hunches, without telling you which is which. The second: they are written for the United States, where there is no right of withdrawal like the European one, so some of them casually recommend restocking fees that, applied to a withdrawal, can end in a penalty here. This guide does the opposite: it starts from why people return according to a broad and recent survey, assigns a tactic to each cause, cites the study measuring its effect where one exists — for three of the seven — and marks where the legal limit lies. If you do not yet know your real rate, start with how to measure the return rate; and if what you want is to recover revenue from the returns that do happen, that is covered by our retained revenue guide.
Why do people really return?
Above all because of size, because products arrive damaged or defective, and because product pages promise something different from what arrives. That is what the most citable survey on reasons we know of says: the one by Bitkom Research for the German digital industry association, carried out among 1,050 online shoppers aged 16 or over in autumn 2023 and published in April 2024. The question allowed several answers, so the percentages do not add up to 100:
| Return reason (multiple answers) | % of shoppers who cite it | Tactic that tackles it |
|---|---|---|
| Wrong size | 67% | 1 |
| Damaged or defective product | 56% | 4 |
| Did not like it | 50% | 2 and 3 |
| Did not match the image or the description | 41% | 2 and 3 |
| Poor workmanship | 37% | 3 and 4 |
| Wrong item sent | 29% | 4 |
| Over-ordered on purpose (several sizes or colours) | 29% | 1 and 6 |
| Made a mistake when ordering | 14% | 1 |
| Late delivery | 13% | 5 |
Three caveats before you use the table. It is Germany, a market with a strong returning culture. It measures how many people have ever cited each reason, not what share of returns corresponds to each one. And it is self-reported: "did not like it" sometimes hides a size problem the customer does not want to explain. Even so, the signal is clear: size rules.
The useful reading is a different one: almost every reason in the table is decided before the parcel leaves. On the product page, in the warehouse or in the delivery promise. That is why prevention is the only lever that reduces returns without touching any right.
Which returns can you reduce without legal risk?
You can reduce the causes of any return, but you can only change the rules of commercial returns, never those of withdrawal. If your shop directs its activity at EU consumers, the customer has at least 14 days to withdraw without giving a reason (Article 9 of Directive 2011/83/EU), except for the contracts Article 16 excludes — personalised goods, perishables or sealed hygiene goods that have been unsealed, among others; we cover them in withdrawal right exceptions. Where that right exists, from 19 June 2026 the customer must be able to exercise it through a clearly identified withdrawal function. What you offer beyond that — longer periods, exchanges, reasons the law does not cover — is your commercial policy, and you design it. The full difference is in withdrawal vs. return vs. exchange.
The practical consequence fits in one rule: work on everything that happens before "buy" and do not put friction into what happens after "I want to return this". If a tactic only works because the customer gives up halfway through the process, you are not reducing returns: you are putting a dark pattern on top of a right.
Tactic 1 — Solve size before checkout
Size is the top return reason, and you tackle it by giving the customer the information to get it right first time. In the Bitkom survey, 67% of shoppers had returned something because of size and 29% had over-ordered on purpose — several sizes of the same item, knowing they would return the ones that did not fit. The two figures are the same problem: fit uncertainty.
What works, from cheapest to most expensive:
- Garment measurements, not size-label measurements. Length, chest width, rise, sleeve length, measured on the actual item. A generic supplier chart does not tell you that these trousers run narrow.
- A fit note per product. "Runs small: if in doubt, size up." It costs one line of text and answers the exact doubt that ends in a return.
- Model with their height and size. "She is 1.72 m and wears an M" turns a decorative photo into information.
- Recommendations drawn from your own returns. If 40% of the returns of a model say "small", you already know what the product page has to say. Your return reasons are a free fit study.
- Size recommenders. They exist as apps; they are worth it mainly in large catalogues with high turnover. Start with the four points above.
In Shopify, per-product measurements are best stored as metafields on the product or the variant, and the fit note in the description or in a block on the theme's product page. The detail by sector — fashion, footwear, sport — is in return policy by industry.
⚖️ Legal limit: ordering several sizes to return the ones that do not fit is legitimate within withdrawal. You can reduce the need to do it; you cannot ban it or penalise it during those 14 days.
Tactic 2 — Make sure the product page does not promise what the parcel does not bring
41% of the shoppers in the survey had returned something because it did not match the image or the description, and another 50% because they simply did not like it. A good part of that "did not like it" is a badly set expectation: a colour that looked different on screen, a fabric that seemed thicker, a size that had no reference in the photo.
What works:
- Photos in neutral light and at least one without colour retouching. If the colour depends on the light, say so.
- Scale in the photo. A hand, a person, a familiar object. Especially for bags, jewellery, décor and furniture.
- Material described as it feels, not as it sells: fabric weight, composition, thickness, whether it is see-through, whether it itches.
- A short video of the product in motion. No production needed: a minute filmed on a phone shows drape, sheen and size better than ten photos.
- What the product is not. "Not waterproof", "stand not included". Clearing it up beforehand stops the customer discovering it when they open the box.
Here the commercial tactic and the legal obligation coincide: the main characteristics of the goods are mandatory pre-contractual information (Article 6(1)(a) of Directive 2011/83/EU). A more faithful product page reduces returns and, along the way, keeps you on the safe side.
Tactic 3 — Let the reviews speak, the bad ones too
Ratings that are too good can push your returns up. That is the finding of Minnema, Bijmolt, Gensler and Wiesel (Journal of Retailing, 2016), with more than two years of data from a large European online retailer — 631,063 purchases of 2,164 electronics and furniture products: overly positive ratings — above the product's historical average — increase purchases, but also returns, because they raise expectations above what the product delivers. The effect is stronger among novice buyers and for cheap products.
The practical conclusion is not "hide the good reviews": it is that the review that reduces returns is the one that informs. Three measures:
- Ask for context in the review: usual size, size bought, height, use. A three-star review saying "runs small, size up" prevents more returns than ten five-star reviews saying "love it".
- Show the distribution of ratings, not just the average, and let people filter the critical ones.
- Reply to negative reviews with information ("this model is narrow in the shoulders"), not with generic apologies.
⚖️ Legal limit: distorting consumer reviews in order to promote a product is on the EU blacklist of unfair commercial practices (Annex I to Directive 2005/29/EC, point 23c, added by Directive (EU) 2019/2161). Selecting only the positive ones can fall under it. Beyond the law, the study gives you an economic reason not to do it.
Tactic 4 — Packaging, quality control and error-free picking
Damaged, defective, poorly made or wrong: these are returns born in your warehouse, not in the customer's decision. In the survey, 56% had returned something damaged or defective, 37% something poorly made and 29% an item they had not ordered. They are also the most expensive returns, because they come with an angry customer.
What works:
- Packaging sized to the product, with filling for fragile items. A big, empty box is a box that knocks things about.
- Sample-based quality control when receiving goods from the supplier, not once they have come back.
- Double check at picking: scanning the barcode of each item against the order or, at the very least, checking SKU and variant before closing the box.
- Record the "damaged" reason with a photo to tell a transport problem from a supplier problem.
A legal nuance worth being clear on: a defective product is not just a return, it is a lack of conformity with its own legal guarantee regime (Directive (EU) 2019/771). Here, reducing returns literally means selling a better product.
Tactic 5 — Promise the delivery time you can meet
A delivery that arrives later than promised increases the likelihood of a return, all the more when you had promised speed. That is what Rao, Rabinovich and Raju (Journal of Operations Management, 2014) found with order and return data from an online retailer: what matters is the consistency between what was promised and what was delivered, and the effect is stronger on orders with express shipping. In the Bitkom survey, 13% of shoppers had returned something because of late delivery.
What works:
- Promise lead times you consistently meet, not those of your best day.
- Warn proactively about a delay before the customer asks.
- Do not sell express if your operation cannot sustain it. According to the same study, a missed promise weighs more when what was promised was express.
The same work brings a second finding worth knowing: orders placed when the shop displayed low stock ("only 2 left!") were more likely to be returned. Urgency drives purchases that do not hold up afterwards. And if that scarcity is not true, it is also illegal: falsely stating that a product will only be available for a very limited time is another of the blacklisted practices (Annex I to Directive 2005/29/EC, point 7).
Tactic 6 — Review your free shipping promotions
Free shipping sells more, but it can push people to buy things that later come back. Two independent studies point to it:
- Shehu, Papies and Neslin (Journal of Marketing Research, 2020): free shipping promotions increase spending on products with a high return risk, which gain weight in the basket and raise the overall return rate. They validated it with a field experiment.
- Lepthien and Clement (Marketing Letters, 2019): in a field experiment with seven shipping fee schedules, free shipping above a minimum order value led to more strategic returns — the customer adds items to reach the threshold and returns them afterwards.
The conclusion is not "drop free shipping": it is measure it net of returns. Compare order value after returns with and without the promotion, not the value in the cart. If your threshold sits just above your average order value, you are probably manufacturing filler returns. Sometimes lowering or raising it is enough for it to stop matching "one more size".
⚖️ Legal limit: you are free to decide how much you charge for outbound shipping. What you cannot do is keep the standard outbound delivery cost when the customer withdraws from the whole order: Article 13(1) of Directive 2011/83/EU requires you to reimburse it. Who pays for the way back is explained in who pays return shipping in the EU.
Tactic 7 — Close the loop: structured reason and outlier products
Without knowing why each item comes back, the previous six tactics are applied blind. The seventh is the one that makes the rest work: collect a structured reason, aggregate it by product and act first on the products that concentrate the most returns.
- A short, closed list of reasons: size, not as described, damaged, arrived late, changed my mind, other. Six options can be analysed; twenty cannot.
- Reason optional for the customer. In a withdrawal you cannot require it — the customer does not have to justify themselves. As voluntary data, it is still very useful.
- Monthly review by product. Sort the catalogue by returns and look at the top ten: they often concentrate a good part of the problem, many times with one dominant reason.
- One action per outlier product. "Small" → fit note. "Not like the photo" → new photo. "Damaged" → packaging. And measure again three months later.
In Shopify, the reason is recorded per order line when you create the return from the admin; if your requests come in through returnEasier, the app does not create that return in Shopify and you have the reason in its auditable export. How to set up the full flow is in how to manage a return in Shopify step by step.
What does not work (or is illegal) even if you read it everywhere
Tightening the rules is the most quoted tactic and the worst documented. The meta-analysis by Janakiraman, Syrdal and Freling (Journal of Retailing, 2016), covering 21 studies, concludes that a lenient returns policy increases purchases more than returns. And in withdrawal, on top of that, most of those rules are not optional:
| Tactic in circulation | Does it work? | Legal in a withdrawal in the EU? |
|---|---|---|
| Flat restocking fee | It may deter; leniency sells more (meta-analysis) | ❌ No: Art. 14(1) of Directive 2011/83/EU only allows passing on direct costs |
| Shortening the return period | In the meta-analysis, a longer period is linked to fewer returns | ❌ Not below 14 days |
| Requiring the original packaging | Deters through friction | ❌ Not as a condition; the route is diminished value under Art. 14(2) if it was over-handled |
| Requiring an account or login to return | Deters through friction | ❌ No: the withdrawal function must be accessible without barriers |
| Hiding the button or the form | Deters through friction | ❌ No: a dark pattern on top of a right |
| Blocking the customer who returns a lot | Does not touch the cause | ❌ Not within the period (BGH, 16-3-2016, VIII ZR 146/15), except in exceptional cases such as fraud |
| Scarcity or countdown messages | Rao et al.: with visible scarcity, more returns | ⚠️ Only if true; false scarcity is on the blacklist |
| Charging for return shipping, informed beforehand | Deters some returns | ✅ Yes, only the direct costs and only if you informed beforehand (Art. 14(1)) |
Only two rows carry no ❌: charging for return shipping, legal in a withdrawal under conditions, and scarcity messages, legal only if they are true — and, even when true, associated with more returns according to Rao et al. Outside withdrawal — in your commercial policy — you can set different fees or periods if you announce them beforehand; on that ground the problem is no longer the law, but the sales it costs you.
Summary: the seven tactics at a glance
| # | Tactic | Cause it tackles (Bitkom 2023) | Evidence | Cost |
|---|---|---|---|---|
| 1 | Solve size before checkout | Size 67%, over-ordering 29% | Cause only (Bitkom) | Low |
| 2 | Faithful product page: photos, scale, material, video | Did not match 41%, did not like 50% | Cause only (Bitkom) | Low-medium |
| 3 | Informative, unfiltered reviews | Inflated expectations | Minnema et al., 2016 | Low |
| 4 | Packaging, quality control and picking | Damaged 56%, poor workmanship 37%, wrong 29% | Cause only (Bitkom) | Medium |
| 5 | Delivery time that is met | Late delivery 13% | Rao et al., 2014 | Medium |
| 6 | Free shipping measured net of returns | Risky and filler purchases | Shehu et al., 2020; Lepthien and Clement, 2019 | Low |
| 7 | Structured reason and outlier products | All | Precondition for measuring the others | Low |
How do you know whether a tactic has worked?
By comparing order cohorts, not months of returns, and ideally against a group that did not get the change. Three rules:
- Measure by 90-day cohorts: one month's orders and how many have come back three months later. Measuring by calendar month while you grow misleads you.
- Isolate the change. If you change the product pages of 20 products, compare their rate before and after with that of 20 similar products you did not touch. Without a control group, any season will prove you right.
- Look at net margin, not just the rate. A tactic that lowers returns and lowers sales even more does not work. Free shipping is the typical example.
And allow for a measurement effect this year: from 19 June 2026 many requests that used to arrive by email come in through a channel that leaves a record. If your measured rate goes up in the second half of the year, do not conclude that your tactics have failed; we go into it in the return rate guide.
What does returnEasier do to help you reduce returns?
It gives you the data that tactic 7 needs without adding friction for the customer. In the legal withdrawal — and, from the Pro plan, in the commercial return too — the customer can choose an optional reason from a short list (changed my mind, wrong size, not as described, damaged, arrived late, other). It is never mandatory: in a withdrawal, moreover, it could not be.
That reason travels in the auditable export, available on every plan, including the free one, together with the type of each request. On the Scale plan (79.99 USD/month) you also get the analytics dashboard and configurable alerts: you can ask to be notified when a specific product exceeds a percentage of your returns. It is only honest to say what it measures: what share of your returns includes that product, not its rate over units sold, because returnEasier does not obtain the units sold of each product, which is the denominator that would be needed. It serves what matters here, which is finding the product that is costing you the most.
What returnEasier does not do is reduce returns by putting obstacles in the way: the withdrawal button is on every plan, with no login and a fixed label. Prevention happens on your product page and in your warehouse; we tell you where to look.
Common mistakes
- Applying tactics without knowing the reason. A size guide does not fix a packaging problem.
- Copying the US playbook. Restocking fees and shortened periods do not transfer to a withdrawal in the EU.
- Filtering out negative reviews. Besides the legal risk, they inflate expectations and send the product back to you.
- Measuring the promotion in the cart. Free shipping is assessed on the value that stays after returns.
- Promising best-day deliveries. A delay on a promise of speed is what generates the most returns.
- Making the reason mandatory. In a withdrawal you cannot, and in the commercial flow it gives you worse data: the customer picks anything to move on.
- Declaring victory with one month of data. Returns arrive weeks later; wait for the cohort.
Frequently asked questions
What is the most effective way to reduce returns? Tackling the cause before the purchase. In the Bitkom survey (2023), 67% had returned because of size, 56% because of a damaged or defective product and 41% because it did not match the product page. Measurements, faithful photos, unfiltered reviews, packaging and good picking tackle those causes without touching any right.
Can I reduce returns by tightening my policy? Only on the commercial side, and it probably will not pay off: according to the meta-analysis by Janakiraman, Syrdal and Freling (2016), leniency increases purchases more than returns. The 14-day withdrawal right cannot be cut back.
Can I charge a restocking fee? Not in a withdrawal: Article 14(1) of Directive 2011/83/EU only lets you pass on the direct cost of the return, and only if you informed beforehand. In your commercial policy, yes, if you announce it.
Does free shipping increase returns? It can. Two studies (Shehu et al., 2020; Lepthien and Clement, 2019) found more returns with free shipping promotions and with minimum thresholds. Measure it net of returns before deciding.
Do five-star reviews reduce returns? Not necessarily: Minnema et al. (2016) observed that overly positive ratings raise sales and also returns. The ones that help are those that report on fit, material and defects.
Can I block someone who returns too much? Within the withdrawal period you cannot refuse them the return of a contract already concluded (BGH, VIII ZR 146/15), except in exceptional cases such as fraud. Work on the causes.
Conclusion
If you have a Shopify shop selling to European consumers, this is the order we recommend: first, record a structured reason so you know where your problem is (tactic 7); then, tackle size and the product page on the ten products that come back most (tactics 1 and 2), because it is the cheapest and moves the most; in parallel, review packaging and picking if "damaged" or "wrong item" appear near the top (tactic 4). Delivery promises and shipping promotions (tactics 5 and 6) come after, with data in hand.
And one rule that holds for all seven: if the tactic acts before "buy", go ahead; if it acts after "I want to return this", in the EU it is probably a legal risk dressed up as a saving.
💡 Want to know which product is costing you the most returns? returnEasier collects the reason when the customer wants to give it, without adding friction, and hands it to you in the auditable export from the free plan; per-product alerts come with the Scale plan. Try it free — 3 trial returns, no card.
Sources
- Directive 2011/83/EU on consumer rights (arts. 6, 9, 13 and 14) — EUR-Lex
- Directive (EU) 2023/2673 — EUR-Lex
- Directive 2005/29/EC on unfair commercial practices (consolidated text, Annex I) — EUR-Lex
- Directive (EU) 2019/2161 (Omnibus) — EUR-Lex
- Directive (EU) 2019/771 on the sale of goods — EUR-Lex
- BGH, judgment of 16 March 2016, VIII ZR 146/15 — Bundesgerichtshof
- Bitkom, «Online-Shopping: Jeder zehnte Kauf geht zurück» (17-4-2024)
- Minnema et al., «To Keep or Not to Keep: Effects of Online Customer Reviews on Product Returns», Journal of Retailing (2016)
- Rao, Rabinovich and Raju, «The role of physical distribution services as determinants of product returns in Internet retailing», Journal of Operations Management (2014)
- Shehu, Papies and Neslin, «Free Shipping Promotions and Product Returns», Journal of Marketing Research (2020)
- Lepthien and Clement, «Shipping fee schedules and return behavior», Marketing Letters (2019)
- Janakiraman, Syrdal and Freling, «The Effect of Return Policy Leniency on Consumer Purchase and Return Decisions: A Meta-analytic Review», Journal of Retailing (2016)
Informational content; not legal advice. For specific cases, consult a lawyer specialised in consumer law.