Since 19 July 2026, large companies selling apparel, clothing accessories and footwear in the EU can no longer destroy the stock they fail to sell. Most of the commentary since has asked the same question: what do we do with the leftovers now that disposal is closed off?
It is a reasonable question. It is also the wrong one, or at least the last one. By the time unsold stock is sitting in a warehouse at the end of a season, nearly every decision that determined its fate has already been taken: what it was priced at on day one, how quickly anyone noticed it was not moving, whether the first markdown came in week six or week fourteen, and whether it was ever in the right market to begin with.
The regulation does not create that problem. It makes the cost of it visible, and it removes the cheapest way of making it disappear.
Key takeaways
- The EU ban on destroying unsold apparel, clothing accessories and footwear applies to large companies from 19 July 2026, and to medium-sized companies from 19 July 2030.
- “Destruction” is read broadly. Recycling and energy recovery are treated as disposal routes, not as a workaround.
- Alongside the ban sits an annual disclosure obligation, with a mandatory reporting template from 2 March 2027 and five-year record retention.
- Between 264,000 and 594,000 tonnes of textiles are destroyed in Europe each year, or 4 to 9% of everything placed on the market.
- The commercial upside does not depend on the regulation. McKinsey puts the margin-rate improvement available from better markdown decisions at 4 to 8 percentage points.
- The durable response is not deeper clearance. It is earlier, better-informed pricing decisions across the product lifecycle, so that less inventory ever reaches the point where clearance is the only option left.
Clearance is where the cost surfaces, not where it is created
The scale of the underlying problem is well documented. The European Environment Agency puts the figure at between 264,000 and 594,000 tonnes a year, with unsold stock and returned goods both contributing. Roughly one fifth of unsold textile inventory has historically been destroyed rather than resold.
Notice what that figure represents. It is not a failure of clearance execution. It is the accumulated residue of pricing and assortment decisions taken months earlier, across thousands of individual products, each of which looked defensible in isolation.
This matters commercially as well as environmentally, because the economics of late intervention compound in three separate ways at once.
The first is arithmetic. Remaining weeks of season are the resource a markdown spends. A product with twelve weeks left needs only a modest lift in its weekly rate of sale to clear, and a modest price reduction will usually deliver it. The same product with four weeks left needs to lift its weekly rate several times over, and the reduction required to move demand that far is correspondingly deeper. The later the intervention, the more work each remaining week has to do.
The second is competitive. Late markdowns land in the same window as everyone else’s. A reduction that would have stood out in week five is unremarkable in week thirteen, because the product is now being judged against a market full of discounted alternatives rather than against its own full price.
The third is behavioral, and it is the one that turns a seasonal problem into a structural one. Customers who learn that a retailer’s stock is always heavily discounted by January stop buying it in November. Habitual late clearance does not only cost margin on the units cleared. It erodes full-price sell-through across the assortment, which produces more unsold stock, which requires more clearance.
And whatever still does not move at the end of all that can no longer simply be written off and destroyed.
Getting better at that final step is worth doing. It is just not where the leverage is.
The lifecycle questions that actually decide the outcome
A lifecycle view of pricing asks a connected series of questions, each at the point where the answer can still change something.
Was the launch price right? The first price is the single largest determinant of how much inventory becomes a problem later. Set it too high on a product with elastic demand and the sell-through curve is compromised from week one, no matter how well the season is managed afterwards. Set it too low on a product with inelastic demand and margin is given away that no amount of full-price selling will recover. Launch pricing decided by cost-plus rules or by last year’s equivalent style is a bet placed without looking at the odds.
How is demand actually developing, and when should we intervene? Early sell-through data is the most honest signal a retailer gets, and it arrives long before the team that makes the pricing decision is ready to act on it.
The concept doing the work here is the expected sell-through curve: the share of buy quantity a product should have sold by each week of its season, given its price, its category’s seasonality profile, and how comparable products performed in previous seasons. It is not a target. It is a prediction, and its entire value lies in the gap between it and reality.
That gap is what makes early data usable. A raw sell-through figure means nothing on its own. The same number can describe a product comfortably ahead of plan or one in serious trouble, depending entirely on what its own curve predicted for that week. The comparison carries all of the information: how far from expectation the product sits, in which direction, and how many weeks remain to close the gap. The question is never whether a product is selling. It is whether it is selling in line with what its own curve predicted.
The same size of gap also means different things at different points in the season. A product falling behind in week three and the same product falling behind by the same margin in week eleven are different decisions, because the first still has enough season left for a modest price change to correct it and the second does not. Most markdown calendars ignore this entirely: fixed dates, fixed depths, applied across the assortment. The lifecycle alternative is to treat the intervention point as a decision in its own right, triggered by the gap between actual and expected sell-through against the weeks remaining, not by the date on the wall.
Are competitors discounting, and does it matter? Competitive movement is constant and mostly irrelevant. The discipline is separating the moves that genuinely shift demand from the noise. A rival dropping price on a product your customers cross-shop is a signal. A rival dropping price on something your customers never compare is not, and matching it simply funds someone else’s clearance out of your margin.
Which products need a markdown, how deep, and when? These are three separate questions that are usually answered as one. Product segmentation matters here more than anything: a traffic-driving style, a margin-generating style and a long-tail style should not share a markdown rule, because they do not share a demand curve, a competitive exposure or a strategic purpose.
Could this inventory perform better somewhere else? Price is not always the right lever. Stock that is dead in one market may be seasonally live in another, and a product failing online may perform in stores where it can be seen and tried. Reallocation is a pricing decision too, because the comparison is always between the margin recoverable here after a markdown and the margin recoverable there at a higher price, net of the cost of moving it.
Connected capabilities, not a checklist of features
Read those questions together and something becomes obvious. Demand forecasting, seasonality modeling, price elasticity, competitive sensitivity and product segmentation are usually bought, deployed and evaluated as separate capabilities. In a lifecycle view they are not separable at all.
Elasticity without segmentation produces a single blunt answer applied to products with entirely different roles. Competitive monitoring without elasticity produces reflexive price matching. Forecasting without seasonality mistakes a normal early-season curve for underperformance. Segmentation without sell-through tracking produces a tidy taxonomy that nobody acts on. Each capability is only as useful as the others it is connected to, because each lifecycle decision draws on several of them at once.
That is also why the answer to the new rules cannot be a compliance workstream bolted onto the end of the season. The decisions that determine how much unsold stock exists are taken continuously, across the whole lifecycle, by the pricing and merchandising functions. Compliance follows from those decisions being made well. It cannot be retrofitted after they have been made badly.
The commercial case does not depend on the regulation
It is worth saying plainly: none of this is new advice, and none of it needs a regulation to justify it. Retailers have always lost money on inventory that reached the end of season in the wrong quantity at the wrong price.
What is striking is how thoroughly the industry has normalized it. Markdown budgets are planned for. Terminal stock is provisioned for. A share of the buy is written off before the season has even started. None of it registers as failure, because the defense is always available: everyone else’s numbers look the same. A cost that every competitor carries stops being read as a cost and starts being read as weather. The disclosure obligation is what breaks that. Large companies must now publish, each year, the volume and weight of unsold products they discarded and the reasons why. Published figures are comparable, and a number that sits beside a competitor’s is much harder to accept as the way things are.
What has changed is the cost of the fallback and the visibility of the failure. McKinsey’s State of Fashion 2026 reports that nearly three-quarters of fashion executives expect to raise prices, with 26% planning increases above 5%, while Levi’s and others have publicly described pulling back from habitual promotional activity to protect brand positioning and margin. That is an industry already trying to get off the discount treadmill. The EU rules simply remove the escape hatch at the end of it.
Retailers that treat 19 July 2026 as a disposal logistics problem will spend the next few years building donation pipelines and secondary market relationships for volume they should never have been holding. Retailers that treat it as a pricing problem will build the capability to hold less of it in the first place, and the donation pipeline becomes a genuine last resort rather than an operating dependency. The second route also carries a return the first does not. McKinsey puts the margin-rate improvement available from markdown optimization at 4 to 8 percentage points. Disposal logistics, however well executed, returns nothing.
What this looks like in practice
The shift a lifecycle view asks for is smaller than it sounds and harder than it looks. It is a move from a markdown calendar to a monitoring cadence: instead of asking twice a season what needs discounting, teams ask every week which products have diverged far enough from their expected curve to justify acting, and which have not. Most of the answers are “not”, which is the point. The discipline is as much about leaving prices alone as about changing them.
Retailers can size the opportunity before changing anything. Take last season’s unsold stock, line by line, and ask two questions of each: at what point did the data first show it would not sell through at the current price, and how many weeks passed before the first markdown was taken? The gap between those two dates, multiplied across the assortment, is the cost of late intervention. It is rarely calculated, and it is usually larger than the clearance discount everyone does argue about.
The obstacle is more often organizational than technical. Launch price sits with buying or commercial. In-season demand sits with planning. Markdown authority sits with merchandising. The margin consequence lands with finance, and now the disclosure obligation lands with legal or sustainability. Each function optimizes its own step competently, and nobody owns the sequence. A lifecycle view of pricing requires a shared view of the product’s life, with the same signals in front of all of those teams at the same time. That is a governance decision as much as a systems one, and it is the one that tends to get deferred.
There is a fair objection at this point: these teams are already at capacity. Pricing typically occupies a small share of a merchandiser’s week, behind range planning, supplier management and allocation. Asking them to review the assortment weekly instead of twice a season sounds like adding a job to people who do not have room for one.
It is not what a lifecycle view actually requires. The weekly question is not “what should we do with all of these products”. It is “which of these products has moved far enough from expectation that a person needs to look at it”. Across an assortment of tens of thousands of styles, that is a list of tens, and producing it is work for a system rather than a team. What expands is the judgment applied to genuine exceptions, which is the part merchandisers are good at and the part they rarely have time for. What shrinks is the blanket seasonal review, the spreadsheet reconciliation and the end-of-season markdown scramble, which consume far more of the week than anyone credits and produce worse decisions than the time they cost. Run properly, this takes less of a merchandiser’s week than the calendar it replaces. The technology question only becomes real once the work has been framed that way.
How Quicklizard fits
Quicklizard is an AI-powered dynamic pricing platform built for omnichannel retailers making these decisions at scale, across large assortments, in markets where the right answer changes weekly. It sits alongside existing forecast and planning software rather than replacing it. A planning forecast asks how much of a product to buy and where to place it, largely before the season starts. A pricing forecast asks what happens to demand for that product if its price changes during the season, and what that does to margin and to the weeks of cover remaining. The two are complementary views of the same product, and the pricing view works from the planning forecast rather than competing with it.
The platform brings the lifecycle capabilities together rather than treating them as separate tools: demand forecasting and seasonality modeling that establish what a product should be doing at this point in its life, elasticity measurement that turns “we think it is overpriced” into a quantified expectation, competitor price monitoring paired with demand-impact analysis so teams respond to the competitive moves that actually matter instead of matching everything, and product segmentation so that markdown logic reflects what each product is there to do.
Because those recommendations are explainable and auditable rather than opaque, pricing, merchandising and finance teams can see why a product is flagged for intervention in week five, and can defend the decision afterwards. Under a disclosure regime, that auditability stops being a nice-to-have.
Less inventory reaching the end of the season in trouble is a better commercial outcome. It now happens to be a better regulatory one as well.
Frequently Asked Questions
Why are the EU’s unsold goods rules a pricing issue rather than a waste issue?
Because the volume of unsold stock is set by decisions taken throughout the product lifecycle: launch price, in-season demand response, markdown timing and depth, and inventory placement across markets and channels. Clearance is the last step in that chain, and by then the options are limited and expensive.
How is managing price across the product lifecycle different from markdown optimization?
Markdown optimization decides how deep to cut once a product is already in trouble. Managing price across the lifecycle means asking the question earlier and more often, starting with whether the launch price was right, and treating the markdown as one possible outcome rather than the only lever. The aim is fewer products reaching the point where a markdown is the only option left.
What is an expected sell-through curve?
The share of a product’s buy quantity that should have sold by each week of its season, based on its price, its seasonality profile and the performance of comparable products. Comparing actual sell-through against it is how retailers identify products in trouble while there is still season left to act.
When should a retailer take the first markdown?
When the gap between actual and expected sell-through is large enough that the remaining weeks of season cannot close it at the current price. That point falls at a different week for different products, which is why fixed markdown calendars systematically intervene too late on some products and needlessly early on others.
Does reviewing prices weekly mean discounting more often?
Usually the opposite. A weekly comparison of expected against actual sell-through finds that most products are tracking as predicted and need no action at all. What changes is that the small number genuinely diverging get caught while a modest price change can still correct them, which reduces both the depth and the volume of discounting later in the season.
Does pricing software replace forecast and planning software?
No. The two answer different questions. Forecast and planning systems size the buy and decide where stock goes, mostly before the season begins. Pricing software works out what happens to demand, margin and remaining cover if the price moves once the season is running. It reads from the planning forecast rather than replacing it.
