Returns cost up to 39% of an item's price: how to turn them into a margin lever

Summarize with AI:

Between return shipping, quality control and restocking, processing a return costs 20 to 39% of the item's price on average (Gartner). It is still one of the least equipped parts of the supply chain: handled case by case, rarely measured, rarely connected to the rest of operations. It is also one of the most decisive for what comes next, since 92% of shoppers who go through a smooth return process buy from the same merchant again (UPS).

The point is not to eliminate returns, but to decide what they become: a refund, or a second transaction. That means handling the customer experience and the physical product in the same place, from the return portal through to restocking, the approach we have taken at Bigblue.

A European regulatory framework that tightened in 2026

Two deadlines changed the rules across the EU this year: since 19 June 2026, the right of withdrawal must be accessible in one click; since 19 July 2026, destroying unsold textiles and footwear is banned. Returns have to get simpler for the shopper, and returned products have to be recovered rather than written off. The volume to absorb, meanwhile, depends mostly on the category: a 25% return rate in apparel and 17% in footwear, against 9% in beauty (Statista, Consumer Insights 2025).

On both counts, the answer sits in the tooling rather than in a compliance project. The Bigblue Return Portal already opens a request in one click, with no printer and no email exchange, and our warehouses run a recovery flow that puts the item back into sellable stock instead of writing it off. For the brands we operate, neither deadline called for a change.

The cost of a return is split across four budgets

If that cost is so rarely tracked, it is because it never appears as a line of its own: 69% of the companies surveyed by Gartner do not measure the true cost of their returns. It is the most common blind spot among the brands we work with, and it comes down to how widely the cost is spread.

Margin absorbs the first part: the item comes back late, often discounted, and in fashion a product frequently takes 30 to 45 days to become sellable again, long enough for a season to pass (Optoro / McKinsey). Customer service absorbs the second, in "where is my refund?" tickets. Warehouse operations absorb the third, in manual inspection time. Revenue closes the list, because a slow refund pushes back the next purchase.

Two levers are enough to bring the bill down. Speed first: across the brands Bigblue operates, 48 hours pass between the dock and sellable stock, enough to put the item back on sale before the season discounts it, with a 0.02% processing error rate. Visibility second: the platform tracks the status of every return in real time, merchant side and customer side, and releases the refund as soon as the carrier scans the parcel, without waiting for it to reach the warehouse. The "where is my refund?" tickets drop out of the support queue.

Part of the return flow is decided before delivery even happens

Not every return comes from product dissatisfaction. A significant share of the flow comes from a delivery that failed: wrong address, parcel never collected from a pickup point, recipient unavailable, or parcel refused. These carrier returns sit under no customer return reason, which keeps them largely invisible in brand dashboards.

« Even when customer returns are low, carrier returns are often a hidden cost for brands and remain a significant issue. To reduce them, we focus on making delivery addresses reliable and on notifying customers at every step of the post-purchase journey. »
William Meunier, co-CEO and co-founder of Bigblue

The lever therefore sits upstream, before the parcel ever travels back. Three automations run by default on the orders we ship: delivery address verification at checkout, notifications at every step of the post-purchase journey, and a reminder email when a parcel is waiting at a pickup point. Every delivery that lands first time is return shipping, quality control and restocking avoided.

Keeping the money inside the brand's ecosystem

A refund is a lost sale, where an exchange or store credit creates a second transaction. It is probably the most profitable call a brand can make on its returns.

« Once the return has been made, the challenge is to keep the money tied to that refund inside the brand's ecosystem and turn it into new revenue. »
William Meunier, co-CEO and co-founder of Bigblue

Shopper expectations point the same way: 76% of US shoppers favour a brand offering an instant refund or exchange (NRF / Happy Returns, 2025), and in the UK, 51% say a returns policy has already influenced a purchase decision (YouGov, 2025).

Everything then comes down to the return screen: depending on whether it leads with the refund or with Exchange and Store Credit, the same request ends as a lost sale or as a second order. The Bigblue Return Portal puts Exchange and Store Credit ahead of the refund option, and pairs every return reason with a product recommendation. Across the brands we operate, that shifts the numbers:

« A customer who chooses store credit over a refund spends 34% more on average: with Bigblue, a return is no longer lost revenue, it is a retention lever. »
CAVAL, responsible sneaker brand

A return portal on its own is not enough

Most brands pick between two incomplete options: a return portal, which handles the customer-facing screen but not the product, or a 3PL, which handles the product but not the experience. The outcome is the same either way: a break in the middle of the flow, interfaces to maintain, and data that travels badly.

« On returns, all four of our solutions play a part: the return portal starts the process on the customer side, the platform gives merchants visibility on how returns are handled and on the KPIs, the TMS manages transport, and the WMS takes care of quality control, right through to automating certain decisions. »
William Meunier, co-CEO and co-founder of Bigblue

Built in house and operated together, Bigblue's four building blocks close that gap. Returns run printer-free in more than 20 countries, with home pickup in markets such as Spain and the UK, at a time when 79% of Europeans drop their parcel at a pickup point or locker (DHL). The 12 European warehouses in the network (France, Spain, UK, Germany, Italy) process returns domestically, with country-specific labels. And the Shopify sync pushes returns and refunds back into the merchant's own tools, including when the request started in a third-party portal.

That continuity also lets each brand set its own cursor between speed and security: ship the replacement item straight away, or wait for quality control to rule out any risk of fraud.

« This automation lets us keep processing costs at around €1.50 per return, while giving us full visibility on every request. »
Cabaïa, fashion accessories brand (peaks of 10,000 orders a day)

What AI actually changes on a return

On returns, AI moves the process from a binary decision, refund or restock, to a judgement call. Across the Bigblue suite, it works at four levels today.

It turns the return into a sale first, recommending the right product for the reason given: another colourway if the colour disappoints, another size if the fit is wrong. It then reduces returns at source, analysing reasons and customer comments to turn them into statistics the brand can act on.

« If 70% of products come back because they are too small, it is worth reworking the size guide. »
William Meunier, co-CEO and co-founder of Bigblue

AI also speeds up and de-risks quality control, with our WMS guiding the operator at each step while computer vision flags product defects. And it optimises the return journey, picking the best carrier by postcode from a model trained on more than 10 million parcels. The next building blocks are already in progress, from fraud detection to refund automation, with one limit we are upfront about: badly calibrated, AI turns down a legitimate refund or over-inspects certain customers. Its job is to support the operator, not to replace them.

« The goal is to turn the return into a new sales opportunity. »
William Meunier, co-CEO and co-founder of Bigblue

A returned product is not a lost product

A large share of returns come back untouched, or with nothing worse than damaged packaging, most often a shipping label stuck straight onto the product box. With the European ban on destroying unsold textiles and footwear, recovery has moved from good practice to obligation.

Three operations make it systematic in our warehouses. Refurbishment first: repacking, relabelling and barcode reprinting on the spot, to put the item back into sellable stock. Sorting next: on arrival, our returns stations, equipped with photo evidence, UV lamps and cameras, establish the real condition of each product and route it to its best destination, from standard stock to clearance and private sales, through resale and, as a last resort, recycling. Documentation last, which makes it possible to justify the decision taken on each product under traceability requirements. A chain set to become a little more automated at every step.

« Tomorrow, it may be enough to place a product under a camera for it to determine automatically whether it is fit for resale when it comes back. »
William Meunier, co-CEO and co-founder of Bigblue

The return as the starting point of the next sale

The best performing brands on this topic are not the ones recording the fewest returns, but the ones processing them quickly, analysing them and converting them back into revenue.

The two sides of the subject, the customer experience and the physical product, are best not separated: return portal, transport, warehouse, restocking and recovery operated as one. That integration is what allows the brands Bigblue operates to reach 48 hours from dock to sellable stock, 96% buyer satisfaction and up to 47% of value recovered per return.