Turning Undeliverable Parcels Into Revenue: The Shipment Exception Lesson From Vinted Go's 100,000-Parcel Experiment

Every e-commerce shipper eventually meets the same parcel. It left the warehouse on time, it was scanned by the carrier, and then something went wrong. The recipient never answered the door, the delivery point closed before pickup, or a label was damaged and the item was sent back with no clear next step. Most businesses have a name for what happens after that: write it off. Vinted Go, the logistics arm of the European resale platform Vinted, decided to test whether that assumption was actually true. Through a program called Rescued Value, the company began relisting undeliverable and unreturnable parcels for sale under a dedicated account instead of sending them to disposal. As of late July 2026, more than 100,000 packages had found a second buyer through the program, available to shoppers across France, the Netherlands, Belgium, Italy, Spain, Portugal, Ireland, Germany, Austria and Luxembourg.

On the surface this looks like a resale platform doing what resale platforms do best, extending the life of secondhand goods. Look closer and it is a logistics story, not a marketplace story. The interesting part is not that Vinted Go found buyers for orphaned parcels. It is that the company built the operational machinery to identify which parcels were truly undeliverable, separate them from parcels that were merely delayed, route them into a new sales channel, refund the original buyer, compensate the original seller, and still come out ahead financially. That is a shipment exception management problem wearing a resale wrapper.

What actually happens to a “failed” parcel

Ask most e-commerce operators what happens to an undeliverable parcel and the honest answer is often that it depends who notices first. A parcel can end up undeliverable for a long list of reasons: a recipient who never collects it from a locker or service point, a shipping label the carrier cannot scan, an address that turns out to be wrong, a customer who refuses delivery, or a return that never makes it back to the seller because it was lost, damaged or abandoned somewhere in the reverse logistics chain. Vinted Go's own account of the program makes clear these are not rare edge cases. They are a steady, predictable stream of volume, and the default handling model across most of the logistics industry is disposal, discounted bulk resale to liquidators, or simply absorbing the loss.

The cost nobody puts on the income statement

The financial case for handling this differently is easy to underestimate, because the cost of a failed delivery rarely shows up as a single line item. It gets scattered across customer support tickets, reshipment costs, carrier surcharges, write-offs, and customers who quietly stop ordering. Independent research into last-mile delivery has repeatedly put the direct cost of a single failed delivery attempt in Europe at somewhere around 14 euros once a redelivery or return leg is triggered, and estimates that close to one in five e-commerce shipments experiences some kind of delivery exception on the first attempt. Multiply that across a mid-sized shipper moving tens of thousands of parcels a month and the number stops being a rounding error. It becomes a material drag on margin, and for scaling e-commerce brands, margin is usually the scarcest resource in the business.

Vinted Go has been explicit that giving these parcels a second life is operationally more complex and more expensive than simply disposing of them. Someone has to inspect the item, decide whether it is sellable, relist it, price it, and ship it again, this time successfully. The company's own reporting on the program suggests the resale revenue from the Rescued Value account is what makes the economics work, covering the cost of collection, processing and reselling rather than stacking a new line of pure loss on top of an existing one.

Why this matters beyond one marketplace

It would be easy to read the Vinted Go story as a quirk of the secondhand fashion business, where every item already has a resale value baked in by design. That framing misses the real lesson. The program is proof of a broader operating principle: an exception parcel is not automatically a write-off, it is an unclassified asset sitting in the wrong workflow. Whether the asset is a returned t-shirt, an electronics accessory that bounced off three delivery attempts, or a grocery box that missed its delivery window, the same three capabilities decide whether it becomes recovered value or dead cost: how fast you detect that something has gone wrong, how automatically you can reroute it to the right next step, and how cleanly you can reconcile the outcome with the customer, the carrier and your own books.

The infrastructure question nobody asks out loud

Most e-commerce brands do not have a resale marketplace sitting inside their own business, and they should not need one to benefit from this lesson. What they need is the infrastructure layer that Vinted Go effectively built for itself: a system that treats shipment exceptions as a managed workflow instead of a support inbox problem. In practice that breaks down into three distinct capabilities.

Detection, in real time, not after a complaint

The parcels that get recovered are the ones flagged early, while there is still a decision to make. A parcel three delivery attempts deep, sitting in a depot for a week, or returned to a hub with a damaged label is not information that should surface for the first time when a customer emails asking where their order is. It needs to surface automatically, the moment the carrier's tracking event signals a problem, so a rule engine or a person can act on it while the item still has value.

Routing, without a human rebuilding the decision tree every time

Once an exception is flagged, someone has to decide what happens next: redeliver, reroute to a locker, refund, restock, or in Vinted Go's case, relist. Doing that manually does not scale past a few dozen parcels a day. It has to be rules based, tied to carrier, product category, order value and customer history, so the majority of exceptions resolve themselves and only the genuinely ambiguous cases reach a person.

Reconciliation, so the numbers actually close

The part that is easy to skip and expensive to skip is the accounting: making sure the refund to the original buyer, the compensation to the original seller, the carrier claim, and the new sale, if there is one, all tie back to the same shipment record. Without that, a business cannot actually tell whether its exception handling is profitable or just busy. We have written before about what happens to parcels that fall into this exact gap and never get a clean resolution.

How Zineps approaches this as the Operating System for Shipments

This is precisely the layer Zineps was built to sit on top of. As the Operating System for Shipments for e-commerce brands, marketplaces, and the carriers and fulfillment partners they work with, Zineps unifies tracking data, delivery events and exception signals from every carrier into one operational view instead of a dozen disconnected portals. When a shipment enters a state that looks like a Vinted Go style exception, whether that is a failed delivery attempt, a stalled scan, or a return that never reached the warehouse, Zineps surfaces it the moment the signal exists, applies the routing rules a brand defines for its own business, and keeps a clean record of what happened and why.

A fashion resale platform can use that foundation to decide what gets relisted. A direct-to-consumer electronics brand can use the same foundation to decide what gets redelivered automatically versus refunded outright. A marketplace working with multiple fulfillment providers can use it to hold every logistics partner to the same recovery standard instead of negotiating exception handling separately with each one. The mechanism is identical even though the end action differs, which is what makes shipment exception management a genuine logistics infrastructure category rather than a feature specific to one resale platform.

A practical starting playbook

1. Pull the last 90 days of carrier exception codes for every shipment that did not deliver on the first attempt, and quantify what actually happened to each one: redelivered, returned, refunded, or simply written off with no record.
2. Separate temporarily delayed from genuinely undeliverable. Most systems treat both the same way, which means recoverable parcels sit in the same queue as parcels that will resolve themselves within a day.
3. Set explicit, automated rules for the top three or four exception types by volume, rather than routing everything to a manual review queue.
4. Decide, deliberately, what your business's equivalent of relist is. It might be an outlet channel, a restock-and-resell flow, a donation partner, or simply a faster refund that protects the customer relationship. The point is to make it a defined workflow, not an improvised one.
5. Track the recovery rate and the cost per resolved exception as an ongoing operational metric, the same way you would track first-attempt delivery success. What gets measured tends to get funded.

The bottom line

Vinted Go's headline number, more than 100,000 parcels given a second life, is a good marketing story. The more useful number for anyone running e-commerce logistics is the underlying one: how many of your own shipment exceptions are quietly absorbed as cost today that could, with the right detection, routing and reconciliation infrastructure, be resolved, recovered, or at minimum understood well enough to stop repeating. As delivery volumes grow and carrier networks become more fragmented across Europe, that infrastructure question is not going away. Shippers who treat exception handling as a strategic capability rather than a support cost will be the ones who feel the least pain when it does.

If you want to see what shipment exception detection and automated recovery routing looks like on your own carrier mix, talk to the Zineps team. We can show you what it looks like inside a live account, mapped to your own carriers, not a slide deck.