Wallapop Bets on Real-Time AI Discovery

Wallapop, the dominant Spanish C2C marketplace, is integrating real-time AI discovery from Swiss firm Albatross. The move signals a significant departure from traditional, history-based recommendation engines towards adaptive, in-session personalization.

This strategic partnership positions Wallapop as an early adopter of dynamic AI discovery in the consumer-to-consumer space. The technology aims to understand user intent as it evolves during a browsing session, offering a more intuitive and responsive shopping experience.

Beyond Static Recommendations

Instead of merely showing more of what a user has already viewed, Albatross's AI adapts to changing interests within a single session. For instance, a user browsing sofas might be shown complementary items like rugs or lamps if the AI detects a shift towards furnishing a room, even highlighting products the user hadn't explicitly searched for.

This approach allows for the discovery of highly relevant, unexpected items. As user behavior changes dynamically, the AI adjusts its recommendations instantly, ensuring listings surface at the optimal moment. This is particularly crucial for recommerce, where unique items can easily get lost.

Quantifiable Impact

Albatross's technology, recently bolstered by a $12.25 million funding round, focuses on understanding user needs in the moment. Forbes has dubbed this approach the “second pillar of AI,” complementing generative AI’s content creation capabilities.

Initial A/B tests on 10% of Wallapop’s traffic demonstrated substantial improvements: an 118.9% increase in user engagement, a 104.8% rise in favorites and interactions, and a 46.9% increase in purchase intentions. Crucially, the system surfaced previously unseen items, indicating its ability to unlock latent supply rather than just promote popular goods.

These gains have proven stable over a four-month production period, validating the robustness of real-time AI discovery. This capability is vital for AI discovery in recommerce platforms.

Seller Visibility and Catalog Challenges

For Wallapop, the partnership extends beyond buyer relevance to algorithmic seller discovery. The system dynamically matches buyer intent with available listings, addressing the challenge of ensuring quality supply isn't overwhelmed by sheer volume in second-hand markets.

"Our collaboration with Albatross represents another step forward in our mission to empower people to participate in a more conscious consumption model," said Rob Cassedy, CEO of Wallapop. "We are moving toward a system that understands what users want in real time, helping buyers find the right items faster while giving sellers more effective visibility."

Albatross's AI continuously interprets live user interactions, updating its understanding of intent and the catalog in real time. This contrasts with conventional recommenders reliant on past behavior and popularity signals. The dynamic nature is ideal for marketplaces with millions of unique, inconsistently-tagged second-hand items.

Dr. Kevin Kahn, CEO of Albatross AI, stated, "Together with Wallapop, we’re delivering an online discovery experience comparable to the best in-store journey." He highlighted that while large language models process words, Albatross's model interprets the sequence and context of user actions. This partnership marks a significant shift in AI discovery, an area with limited innovation for over a decade.

The Wallapop Albatross partnership is one of the earliest large-scale commercial deployments of adaptive, in-session AI discovery in recommerce, reinforcing Wallapop’s leadership ambitions. This innovation echoes advancements seen in other AI discovery initiatives.