Why a fashion webshop at 40 percent returns no longer makes anything on an extra order

The third-quarter figures look good. The number of orders is 18 percent higher than last year and the gross margin is 52 percent. Conversion has also risen since the free returns campaign. Profit has nevertheless fallen. The owner points to advertising costs, which have become more expensive per click again, and to the carrier, which raised its rates in January. Both costs have indeed risen, and together they explain a fraction of the fall. The rest lies in the 40 percent of parcels that come back.

Calculate this before the fourth quarter starts. With fashion webshops we see that December brings the revenue peak of the year and January the returns and refunds. On top of that, on 23 July 2026 (assessment, not a formal decision) the Netherlands Authority for Consumers and Markets (ACM) announced that clothing webshops may cooperate through Thuiswinkel.org on a study into the effect of return costs on buying and return behaviour, on condition that participation is voluntary and each webshop decides for itself what amount it charges. Anyone adjusting their returns policy this autumn is therefore doing so in a market where competitors are considering the same. And according to the study delivered by Rebel, TAUW and Ecolaw (research report commissioned by the Ministries of Infrastructure and Water Management and of Economic Affairs, 2025), 71 percent of Dutch fashion webshops charged nothing for a return at that time, processing a single return costs €12 to €17, and the researchers take as their starting point that about 35 percent of clothing items ordered online are sent back.

What a return really costs a webshop: the worked example

Everything in this section is a worked example with deliberately ordinary numbers: no extreme shipping rates and no 30 percent discount code. The point is the mechanism, which you can then apply to your own figures.

The business is a women's fashion webshop with 30 employees. It processes 100,000 orders a year with an average order value of €96.80 including VAT, €80 excluding VAT. The cost of goods is 45 percent of the selling price, €36, so the gross margin per order is €44. Shipping costs €5.50 per parcel and is free for the customer above €50; picking, packing and packaging cost €2.50; the payment provider charges €1 per order, rounded. The webshop spends €1.2 million a year on advertising, affiliates and email marketing, €12 per order placed. Returns are free: the webshop pays the €5.50 return label and spends €4 per return on checking, repacking and refunding. Of the returned items, 10 percent can no longer be sold at full price; in the example these are written down to zero, an average of €3.60 per return. For comparison: the report by Rebel and colleagues (2025) states that about 90 percent of returned clothing is resold, and that processing a return costs €12 to €17. The €5.50 plus €4 in the example is below that, because outbound shipping and packing are shown separately here.

To keep things clear, the example calculates per order, and a return is a full return. In reality, often one item from an order of three goes back; the mechanism is then the same, only the amounts per line differ.

A kept order delivers €23. A returned order costs €34.10, and the largest part of that, the €12 of marketing, has already been spent by the time the customer clicks to order. The expected contribution per order placed is then 60 percent of €23, minus 40 percent of €34.10. What the kept orders deliver and what the returns cost almost balance out at 40 percent: 16 cents remains. Over 100,000 orders that is €16,000 of contribution, and the fixed costs of office, customer service, rent and software still have to come out of that.

As a formula, with three numbers you take from your own figures: B is the contribution of a kept order, V is the loss on a returned order, R is the return rate written as a number (40 percent = 0.40).

That 40 percent is close to what the industry measures. Thuiswinkel.org reported in 2023 (knowledge article based on 9 million shipped and returned items, 2020 data from the Cross-Border E-commerce expert group) an average return rate of 40.6 percent for women's fashion in the Netherlands, against 50.9 percent in Germany and 19.2 percent in France, and reported that processing a return requires four to five times as much handling as shipping an order. A Dutch women's fashion webshop with free shipping, free returns and €12 of marketing per order is therefore on average running at the limit.

Why you do not see this in the annual accounts

The €34.10 of a return does not appear anywhere as a single amount. Outbound shipping and the return label are in the logistics line, packing and processing in staff costs or fulfilment, the write-down in the cost of goods sold, and marketing in the marketing line. Each line on its own looks like a percentage that fits the industry. The revenue in the annual accounts is also revenue after returns, €4.8 million in the example, and the gross margin on it is 52 percent. That percentage is correct. Whether the last 10,000 orders made money, you cannot tell from it.

With webshops we see that reporting stops at gross margin and the marketing budget as a percentage of revenue. The contribution per order placed, after returns, is rarely calculated, because it requires bringing together figures from four systems: the order platform, the carrier, the payment provider and the advertising account. The head of sales steers on conversion and orders, the marketer on cost per order. The logistics manager looks at the cost per parcel and counts a return as one more parcel. All three meet their targets, and the 16 cents per order appears in none of their reports.

What free returns do to orders and to returns

That the free returns campaign delivered more orders was to be expected. Wood (Journal of Marketing Research, 2001) showed in three experiments that a lenient returns policy lowers the threshold to order and at the same time increases returns less than retailers expect, because once a product is in the home the customer starts to see it as their own and then values it more than before ordering. For a webshop, this means that the customer's wardrobe keeps the returns bin partly empty.

Janakiraman, Syrdal and Freling (Journal of Retailing, 2016) added something in a meta-analysis of 21 studies that matters directly for the calculation above. They split returns policy into five dimensions: time, money, effort, scope and exchange. Leniency on money and effort mainly increases the likelihood of a purchase. Leniency on scope, that is which products may be sent back and in what condition, mainly increases the number of returns, while a longer return period and an exchange option actually reduce the number of returns. Orders and returns therefore respond to different parts of the policy. Anyone who relaxes money, effort and scope at the same time gets both effects, and the bill for that is in the table above.

Petersen and Kumar (Journal of Marketing, 2009) found at a large retailer that customers who occasionally send something back go on to buy more than customers who never return anything, and that this relationship reverses at high return behaviour: above a certain level, each further return costs future revenue instead of generating it. Returns are therefore part of a good customer relationship, up to a limit that differs per customer.

How much customers value free returns is less than most webshops assume. Shang, Pekgün, Ferguson and Galbreth (Journal of Operations Management, 2017) calculated from eBay auctions what buyers pay extra for a full refund on return, and found that this value is smaller than thought, differs between experienced and inexperienced buyers, and practically evaporates as soon as there are outbound shipping costs that are not refunded. Whether a return charge drives customers away is therefore not settled; it depends on the whole package of shipping, refunding and returning.

Calculate per category and per customer: the spread around 40 percent

The 40 percent in the worked example is an average across the whole shop. Money is made or lost per category, and that is where the spread lies.

The percentages per category are chosen for this example; the proportions match what the Thuiswinkel study (2023) names as the main cause of returns: for 70 percent of returned women's clothing, size or fit is the reason, and fit matters more for trousers than for a scarf. With the same €23 contribution and €34.10 loss per return, the webshop loses €8.41 on every trousers order and earns €17.29 on every accessories order. The average of 16 cents hides a loss of €252,000 on trousers against a profit of €173,000 on accessories, and in the worked example the trousers campaign is precisely the campaign that brings in the most orders.

One level deeper is the customer. Buldeo Rai and colleagues (Vrije Universiteit Brussel, 2026, research among almost 10,000 online consumers in ten European countries, published in Transportation Research) found that 15 percent of consumers send back more than half of their orders, on average fifteen products a year, and that this group causes almost 60 percent of the emissions from returns. These buyers order more impulsively and more often order several variants at once. In the worked example, a customer who returns 70 percent of their orders would cost the webshop €16.97 per order. Petersen and Kumar (2009) show that you should not simply turn that customer away, because up to a certain level returning is part of being a good customer. But the contribution per customer, after returns, belongs in the same list as revenue per customer. The article on the Pareto principle as a risk gauge describes how to build such a list.

What return rate is acceptable for a fashion webshop?

An industry norm for the return rate does not help you, because the limit follows from your own contribution and your own return costs. So answer these four questions; they are written so that you can answer them without connected systems, using an export from the order platform and the invoices from your carrier and advertising platform.

1. What is the contribution of a kept order? Gross margin minus shipping, packing, payment costs and marketing per order placed. For marketing, take the whole budget for the last twelve months divided by all orders placed, including those returned. In the worked example €23. If you cannot do it per category, do it for the ten best-selling items.
2. What does a return cost? Outbound shipping, return label, packing, processing, payment costs, write-down and the marketing that brought in the order, added together. In the worked example €34.10. If you have never measured the returns team's time, use the €12 to €17 processing cost from the report by Rebel and colleagues (2025) as a starting point and then measure for a week.
3. Where is your limit? Divide the contribution from question 1 by the sum of questions 1 and 2: B / (B + V). In the worked example 23 / 57.10, so 40.3 percent. If your actual return rate is more than five percentage points below that, every extra order makes money and free returns are a good investment. If it is above, every order the campaign brings in on top costs money, however good the conversion.
4. Which categories and which customers are above that limit? A return rate of 55 percent on trousers, with a contribution of €23, is a loss of €8.41 per order; only at a contribution of €45 per kept order does it become a small profit at the same return costs.

If you are below the limit on all four questions, there is nothing wrong, even at 35 percent returns, and this article is a confirmation for you. That happens, especially with webshops that have their own brand and a gross margin above 65 percent, and with shops whose marketing is largely fixed. In the latter case marketing does not belong in the loss per return: the contribution if kept then becomes €35, the loss on a return €22.10 and the limit 61 percent. The question you then have to answer honestly is whether that marketing budget would really stay the same if 10,000 fewer orders came in. With advertising paid per click or per order, the budget falls along with the orders, and then the €12 belongs in the calculation.

When this story does not apply

The 40 percent limit depends on three assumptions from the worked example: a gross margin of 55 percent, marketing as a variable cost of €12 per order, and returns coming back as a whole order. If any of the three is different for you, you get a different limit.

A webshop with its own brand and a cost of goods of 30 percent has a contribution per kept order of €35 at the same costs. The limit is then 35 / (35 + 34.10) = 0.507, rounded 51 percent, and 40 percent returns is simply profitable for that shop. For a shop with many returning customers who order via email, marketing per order is a fraction of the €12 in the example, and the limit moves in the same direction. And a shop that mainly sells shoes, where the customer orders two sizes and returns one, loses half an order per return and keeps the other half as revenue; that shop needs a table per order line, and the limit is then higher.

The limit can also be lower. If a webshop ships to Germany for €7 per parcel, out and back, where Thuiswinkel.org (2023) reports a return rate of almost 51 percent for women's fashion, the contribution per order in the worked example becomes a loss of more than €8. Anyone wanting to grow in that market first has to adjust the returns policy in this example; otherwise a bigger advertising budget only increases the loss.

What can a fashion webshop do if returns eat up the contribution per order?

Five measures, each with the number from the worked example or the source it relies on. The first is one we earn from ourselves, because setting up management information is our work; so read it with the same scepticism as a quotation.

1. Put the contribution per order after returns on the dashboard, per category and per month. In the worked example, the picture changes from "gross margin 52 percent, fine" to "16 cents per order, trousers lose €8.41". It does not have to be a data warehouse: a monthly export from the order platform with the category, return status and marketing source per order, linked to the carrier's invoice and the advertising account, is enough to start.
2. Charge a return fee, and keep exchanges free. Rebel and colleagues (2025) assumed in their model, based on earlier research and interviews, 15 percent fewer returns with a mandatory return fee, and reported that webshops that already charge something ask €2.26 on average, with a range from 50 cents to just over €5. In the worked example, with a fee of €3.95 the return rate falls from 40 to 34 percent and the loss per return drops to €30.15; the contribution per order then becomes €4.93. Even if the number of orders falls by 10 percent as a result, the annual contribution goes from €16,000 to more than €440,000. If returns unexpectedly stay at 40 percent, the contribution is still €1.74 per order. Janakiraman and colleagues (2016) show that an exchange option reduces the number of returns; free exchanges alongside a paid return combine both effects. On 30 June 2026 (news item), the Dutch Consumers' Association (Consumentenbond) wrote that even a small fee already reduces the number of returns, and called return costs a supplementary measure, after better product information.
3. Solve fit before the order is placed. For 70 percent of returned women's clothing, size or fit is the reason, according to Thuiswinkel.org (2023). Size advice based on previous purchases, a size chart per brand instead of one chart for the whole shop, photos on models of different sizes and a mandatory return reason per item that you review monthly: each of these lowers the return rate without touching the policy. If the return rate in the worked example falls from 40 to 36 percent as a result, the contribution per order rises from 16 cents to €2.44, more than €244,000 a year and €228,000 more than now. The report by Rebel and colleagues (2025) names better fit information and standardised sizing as the measures with the most potential.
4. Steer advertising on contribution after returns, instead of on cost per order. In the worked example, a trousers campaign brings in orders that lose €8.41, and an accessories campaign orders that deliver €17.29. A campaign judged on cost per order does not see that difference. Give the advertising platform the contribution per category after returns as a target value, or keep it simple and lower the bid on the categories above your limit.
5. Decide per customer. Petersen and Kumar (2009) show that returning up to a certain level is part of being a good customer and reverses above it. A list with each customer's revenue, return rate and contribution over the last twelve months shows for whom free returns are a good investment and for whom they are not. What you then do with that last group is a commercial choice; the worked example gives the number: a loss of €16.97 per order at 70 percent returns. Buldeo Rai and colleagues (2026) advise not using a lenient returns policy as a lure and above all giving this group better size and fit information, because according to the study they are more likely to be poorly informed than indifferent.

Whatever you choose here, the fourth-quarter returns come back in January. The money then goes back to the customer within 14 days. The stock comes back in a month in which it will only sell in the sales, and you offset the VAT on the credit notes in the January return. That is how October's returns policy feeds through into January's cash.

What the law says about return costs, and which questions to ask your adviser

Under the rules summarised by the Autoriteit Consument en Markt (Netherlands Authority for Consumers and Markets; explanation for sellers, consulted September 2026), a consumer who buys online has a cooling-off period of 14 days after receiving the product. Within that period they may cancel the purchase without giving a reason. The webshop then refunds everything the customer paid within 14 days, including the delivery costs of the outbound shipment, and may wait until the product is back or the customer has sent proof of shipment; if the customer chose a more expensive delivery option, only the standard option has to be refunded. The cost of sending the item back may be charged to the customer, on condition that the webshop clearly stated this before the purchase; if it did not, those costs are for the webshop. A customer may examine the product as they would in a shop, which for clothing means trying it on; if the product is used beyond that or damaged, the webshop may charge for the reduction in value. According to the same ACM explanation, since June 2026 a webshop must also offer a clearly visible function with which the customer can cancel the purchase within the cooling-off period, a withdrawal button, and inform the customer about it in advance. The study by Rebel and colleagues (2025) concluded that a legally mandatory return fee is not feasible in the short term because of European consumer law; the choice therefore lies with the webshop itself.

We do not give legal advice, and the above shows what you need advice on. The questions for your adviser: how do you word the return fee in the general terms and conditions and in the ordering process so that the information requirement is met; may you charge a return fee for an exchange or only for a refund; which rules apply to your customers in Germany, Belgium and France if you sell there; how do you set out the reduction in value without conflicting with the right of withdrawal; does your ordering process comply with the mandatory withdrawal button and what are the consequences if the information about it is missing; and how do you process the return fee and the credit notes in the VAT return.

Frequently asked questions about returns and the contribution per order at a fashion webshop

How much does a return really cost my webshop?

More than the return label. In the worked example in this article, a returned order costs €34.10: outbound shipping of €5.50, the return label of €5.50, packing €2.50, processing €4, payment costs €1, write-down €3.60 and the €12 of marketing that brought in the order. The report by Rebel and colleagues (2025) puts processing a return alone at €12 to €17, and Thuiswinkel.org (2023) reports that a return requires four to five times as much handling as a shipment.

At what return rate do I stop earning anything on an order?

Divide the contribution of a kept order (B) by the sum of that contribution and the loss on a return (B + V). In the worked example: 23 / (23 + 34.10) = 23 / 57.10 = 0.403, so 40.3 percent. With a higher gross margin or lower marketing costs per order, that limit moves up; for an own brand with a 70 percent gross margin it is around 51 percent in the example.

Am I allowed to charge return costs as a webshop?

According to the explanation from the Autoriteit Consument en Markt (consulted September 2026), the cost of sending items back may be charged to the customer, if the webshop clearly stated this before the purchase. The 14-day cooling-off period still applies, and the webshop refunds the purchase price and the cost of standard outbound delivery within 14 days. Since 23 July 2026, clothing webshops may also cooperate through Thuiswinkel.org on a study into the effect of return costs, with each webshop setting its own amount. Ask your adviser how to set this out in your terms and your ordering process.

What happens to my revenue if I introduce return costs?

Nobody knows exactly in advance; the study for which the Netherlands Authority for Consumers and Markets made room in 2026 is meant to measure exactly that. What is known: Rebel and colleagues (2025) assumed in their model 15 percent fewer returns with a mandatory contribution, Janakiraman and colleagues (2016) found that leniency on money mainly affects the likelihood of purchase, and Shang and colleagues (2017) found that customers value free returns less than retailers assume. In the worked example the measure remains profitable even if the number of orders falls by 10 percent.

Why is my webshop growing while my profit is not?

Because the revenue in the annual accounts is revenue after returns, and the costs of a return are spread over four lines: logistics, staff, cost of goods and marketing. Each line on its own looks normal. The contribution per order placed, after returns, shows whether the extra orders make money. In the worked example that is 16 cents per order at 40 percent returns.

How do I calculate the contribution per order after returns?

Take the gross margin of an order and deduct shipping, packing, payment costs and marketing per order placed; that is the contribution if kept, €23 in the example. For a return, add up all costs that do not come back: outbound shipping, return label, packing, processing, payment costs, write-down and marketing; that is the loss on a return, €34.10 in the example. The contribution per order placed is then B × (1 - R), minus V × R, with R as the return rate written as a number (40 percent = 0.40). In the example: 23 × 0.60 = €13.80, minus 34.10 × 0.40 = €13.64, leaving €0.16. Do this per category, because the shop average hides the loss-makers.

Which categories in my webshop lose money because of returns?

The categories where fit determines the purchase. In the worked example, trousers at 55 percent returns lose €8.41 per order and dresses at 45 percent €2.70, while tops at 30 percent deliver €5.87 and accessories at 10 percent €17.29. Thuiswinkel.org (2023) cites size or fit as the reason for 70 percent of returned women's clothing. With an export from the order platform you can fill this table with your own figures per category.

Further reading: The Pareto principle is also a risk gauge for your business and Cash flow as a brake on growth: why SMEs get stuck and how to break through.

← Back to knowledge base