Most brands reading about ESPR reach for the warehouse. More parts, longer, on a shelf somewhere in Europe.
That is the easy half and it is not the obligation. The regulation is built around proving that a named part reached a named customer inside a set number of days, for a product line that may have stopped selling three years ago.
Availability is a stock question. Provability is a data question, and a part on a shelf that nobody can trace to a claim is not compliance, it is inventory.
This is written for warranty-heavy brands in electronics, furniture, DIY and sporting goods, where products are repairable, parts are real, and the aftersales team is about to inherit a legal obligation that was written for the product team.
What ESPR actually is, in one section
The Ecodesign for Sustainable Products Regulation, Regulation (EU) 2024/1781, replaced and widened the old Ecodesign Directive. The directive covered energy-related products. The regulation covers almost everything physical placed on the EU market, with a short list of exclusions.
ESPR itself sets almost no product requirements directly. It is a framework, and the actual obligations arrive through delegated acts adopted product group by product group, each one carrying its own requirements and its own dates.
Those delegated acts can require, among other things, durability, reparability, spare parts availability, maximum delivery times for spare parts, access to repair information, and a Digital Product Passport. The framework also carries the ban on destroying unsold goods, which the unsold goods side of aftersales has already started feeling.
So the honest answer to 'what does ESPR require for spare parts' is that it depends on the product group and the delegated act, and that several of those acts are still being written. That is not a reason to wait, for the reason set out in the next two sections.
What is already binding today
The spare parts obligation is not hypothetical, and this is the part most compliance briefings bury.
Existing ecodesign implementing regulations already impose parts obligations on several product categories. Washing machines, dishwashers, refrigerators, displays and several others already carry requirements to make defined spare parts available for a set number of years after the last unit is placed on the market, and to deliver them within a fixed window measured in working days.
So the template ESPR is extending has been running for years, on real products, with real delivery deadlines. The regulation widens the scope of who it applies to rather than inventing the mechanism.
Alongside it sits the EU right to repair regime, which pushes in the same direction from the consumer side. Repair obligations, repair information access, and pressure to make repair the default rather than the exception. The practical implications for online brands are in EU right to repair for ecommerce and in repair workflows for EU compliance.
Brands selling appliances into homes have the sharpest version of this already, covered in right to repair for home appliances, and the parallel US movement is tracked in US right to repair laws.
What the delegated acts are being written to require
Every one of those four generates a record the brand has to be able to retrieve later. Which part, for which product, ordered by whom, shipped when, arrived when.
That is not a product data question. It is an aftersales transaction log, and it is generated by warranty claims and repair jobs rather than by the PLM.
The clock that matters: delivery time, not availability
Fifteen working days is not a target. It is a countdown that starts without asking the brand whether it is ready.
The countdown begins when the part is ordered, which for most brands means when a claim or a repair job identifies which part is needed. Every day spent working out which part the customer actually needs is a day off the clock, and it is usually the largest single block of time in the whole cycle.
That is the uncomfortable link between compliance and identification. A brand that takes six days to determine which part failed has already spent 40% of its window before the part has left a shelf. Getting that step down is a data problem, and Claimlane's AI Agent, the first AI agent purpose-built for warranty claims and returns, reads the photos, the serial and the fault description to rank the part before an agent opens the case. The AI Agent applies warranty rules per product and supplier and recommends or auto-approves resolutions, which compresses the front of the countdown rather than the back.
The guardrails on that stay firm. Rules in front of the model, human review above a value threshold, an audit trail on every decision and override controls in both directions, per AI agent guardrails. A compliance record produced by an unauditable model is worse than no record.
The back half of the countdown is stock and logistics, and that half is genuinely covered by spare parts inventory management and by demand forecasting. It is the half most brands have already invested in, which is why the front half is where the days are hiding.
Why the parts data lives in the claims record
The parts a brand actually consumes are the ones its own broken products asked for.
That information is produced at the moment a claim is adjudicated: this unit, this failure mode, this part, this date, this outcome. It is not in the ERP, which sees a stock movement without a cause. It is not in the PLM, which sees a bill of materials without a history. It is in the claims record, or it is nowhere.
A brand that runs claims through email and spreadsheets can tell an auditor how many parts it shipped. It cannot tell them which product line consumed them, how long each one took from request to delivery, or whether the 15 day window was met on any individual case. Those are per-claim facts and they only exist if the claim was a record rather than a conversation.
Analytics running off the claim record produce exactly the cut that compliance asks for: parts consumed by SKU, by failure mode, by supplier, with the elapsed time on each. The same cut is what feeds the supplier scorecard and the warranty analytics that make the parts programme commercially defensible rather than just legally required.
The dealer leg, which is where the proof breaks
A dealer who orders a part by email has just removed that part from the brand's compliance record.
This is the gap nobody on the SERP mentions. Obligations attach to the product placed on the market, and for brands selling through retailers, dealers and installers, a large share of parts requests arrive outside the brand's own systems entirely. Phone calls, emails, an account manager's spreadsheet.
Those requests carry no unit, no failure mode and no timestamp anyone can defend. The part ships, the customer is happy, and the brand has no evidence that it met a delivery window it is legally obliged to meet.
Closing it means putting dealers through the same structured intake as consumers, with the same serial and part identification and different commercial terms. That is the hybrid B2C and B2B claims argument arriving through the compliance door, and the operational version is in dealership warranty management and retailer challenges with supplier claims.
A self-service portal that both consumers and dealers submit through is the least glamorous compliance investment a brand can make and the one that actually produces the record.
The Digital Product Passport ties the two together
The Digital Product Passport is the other half of ESPR, and the two halves meet at the part.
The passport carries product identity, materials, repairability and, depending on the delegated act, information about parts and repair. For it to be accurate over a product's life it has to be fed by what actually happened to the product, which means the repair and parts history has to exist somewhere it can be read from.
A passport populated at manufacture and never updated is a snapshot. A passport connected to the claims and repair record reflects the unit as it exists now. The mechanics are covered in the Digital Product Passport explained, and the identity layer underneath it is serial number tracking.
The adjacent regimes push the same way. WEEE compliance, textile EPR, the EU battery regulation and GPSR all ask a brand to say something specific about a specific unit after it has left the warehouse. Brands answering each one with a separate spreadsheet are building the same dataset five times.
What repairs-first does to the numbers
The legal case is only half of it, and the commercial case is the half that survives a budget review.
Repair is usually cheaper than replacement once the full cost of a replacement is counted: the new unit at cost, the freight, the write-off on the returned unit, and the defective product write-off that lands in the ledger. A part costs a fraction of that, and the decision framework is in repair versus replace.
The first-time fix rate is what makes or breaks it. A repair that needs a second part is worse than a replacement, both on cost and on the delivery clock, which is why identification quality and compliance are the same project wearing two hats.
The finance-readable line for a CFO is cost per claim fully loaded against parts consumed, measured per SKU. Brands that run repairs properly find the parts programme pays for itself before the regulation ever arrives, and the environmental case in the environmental impact of warranty claims is a bonus rather than the argument.
What a brand should be able to produce today
A brand that cannot produce those five lines is not non-compliant yet. It is simply unable to demonstrate compliance, which becomes the same thing on the day someone asks.
The work to fix it is not a compliance project. It is a claims and repair process that happens to leave the right record behind, which is what spare parts management and the underlying workflows are for.
Worth naming the category boundary, as always. Brands whose returns are size and fit on a DTC store do not have a parts obligation and should not buy for one, and the Loop Returns alternatives comparison covers what they should look at instead. Brands with repairable products, real spare parts and a dealer network are the ones ESPR is aimed at, and that is the category Claimlane is built for.
Claimlane is rated 4.8/5 on G2. Cult and Mads Nørgaard are two more brands running claims on one record.
The delegated acts arrive product group by product group, and each one lands with its own date. The brands that will meet them are not the ones that read the regulation first. They are the ones whose repair process was already producing the record.
Building that record takes a season. Reading the delegated act takes an afternoon, and it will happen after the date is already set.
A brand that wants to know where it stands can start with line 02 above: elapsed days from part request to delivery, per case, for the last quarter. Most cannot produce it. Book a demo and bring whatever comes closest.
.webp)
