The secondhand luxury market is now large enough to have an infrastructure problem. The problem is authentication, and it is getting harder, not easier.

Bain and Altagamma estimated the global secondhand luxury goods market at roughly 40 billion euros in 2024, growing at mid-to-high single digits. Platforms like Vestiaire Collective, The RealReal, Vinted's luxury tier, and Rebag are processing hundreds of thousands of items per year. Every one of those items needs to be verified as genuine before it can be sold at a price that reflects its brand.

The fakes have gotten better. Dramatically better. The "super fakes" coming out of certain manufacturing clusters - primarily in southern China, but also Turkey and parts of Southeast Asia - are now so close to the originals that physical inspection by a trained authenticator catches only the most obvious tells. Stitching quality, hardware weight, leather grain, logo placement - the counterfeiters have closed the gap on all of the traditional markers.

What authentication looks like today

The resale platforms currently use a combination of approaches:

Human expert review. Still the backbone. The RealReal employs hundreds of authenticators who physically inspect items. Vestiaire Collective runs a hybrid model with both in-house experts and community-based verification. The cost per item for expert inspection ranges from 5 to 15 euros, depending on the category and the depth of review required.

Computer vision and AI. Most platforms have invested in AI-assisted authentication over the last three years. The models are trained on databases of authentic and counterfeit items and can flag anomalies in photographs - wrong stitch counts, incorrect font spacing on logos, hardware color that doesn't match the reference. These systems work well as a first-pass filter but are not yet reliable enough to replace human judgment on high-value items.

Microscopic and materials analysis. For high-value pieces - handbags over 2,000 euros, watches, fine jewelry - some platforms send items for laboratory-grade analysis. This can include microscopic examination of leather grain, spectroscopic analysis of hardware composition, and comparison against reference samples. It's accurate but expensive and slow.

Digital product passports. Some luxury brands have started embedding NFC chips or other digital identifiers in new products. LVMH's Aura blockchain platform and similar initiatives aim to create a verifiable chain of custody from factory to consumer. In theory, this solves the authentication problem for new items going forward. In practice, adoption is still partial, and the vast majority of items entering the resale market today were produced before these systems existed.

The economics of getting it wrong

The cost of a false positive - authenticating a fake as genuine and selling it - is severe. For the platform, it means a refund, a reputation hit, and potential legal exposure. For the buyer, it means paying thousands of euros for a counterfeit. Several high-profile cases in 2024 and 2025, where platforms sold items later determined to be counterfeit, generated significant media coverage and regulatory attention.

The cost of a false negative - rejecting a genuine item as suspect - is also high, though less visible. Consignors whose authentic items are rejected lose trust in the platform. Given that supply acquisition is the primary bottleneck for most resale platforms, rejecting genuine inventory is a direct hit to growth.

The platforms are caught between these two risks, and the balance is getting harder as fakes improve. Several platform executives we've spoken to describe authentication accuracy in terms that are sobering: 98 to 99 percent is achievable with current methods on most categories, but that still means one to two items per hundred may be misclassified. At the volumes these platforms are now processing, that translates to thousands of potential errors per year.

Where the technology is heading

The next generation of authentication technology is moving in two directions:

Better AI, trained on larger datasets. The platforms are building proprietary datasets of authenticated items - millions of images, measurements, and material profiles - and using them to train increasingly sophisticated models. The goal is to get AI accuracy to the point where human expert review is only needed for edge cases and high-value items. We're probably two to three years from that on mainstream categories like handbags and sneakers, longer for watches and jewelry.

Brand collaboration. The most effective long-term solution is for the luxury brands themselves to participate in authentication. Some are moving in this direction - Richemont (which owns Watchfinder and part of the Farfetch estate) has the infrastructure. LVMH's digital passport initiative is relevant. But most brands remain ambivalent about the resale market. They benefit from the residual value it creates for their products, but they don't want to actively support a channel that cannibalizes new sales.

The margin implication

Authentication cost is a meaningful line in the P&L of every luxury resale platform. At 5 to 15 euros per item on a product with an average selling price of 200 to 400 euros, authentication represents 2 to 7 percent of revenue. That's before you add the cost of errors - refunds, reputational damage, legal.

For the platforms to reach sustainable profitability, authentication costs need to come down while accuracy stays the same or improves. That's the arms race: the technology gets better, but the fakes get better too, and the cost of staying ahead is ongoing.

The platforms that invest most effectively in authentication technology - and that figure out how to collaborate with brands on verification - will have a structural advantage. Authentication accuracy is becoming a competitive moat, not just an operational necessity.