Case Study: Vinted boosts search performance and conversions with Vespa

Vinted cuts search latency 2.5x with Vespa

Vinted, an online marketplace for second-hand goods, faced significant challenges with its existing Elasticsearch implementation as rapid growth led to high operational costs and performance strain. To support its 120 million users and 1 billion searchable items, the company needed a more scalable and efficient solution for its search, recommendation, and personalization engines. They chose to partner with vendor Vespa to address these issues.

By implementing the Vespa search engine, Vinted achieved a massive improvement in performance and efficiency. The solution reduced their server count from 120 to 60, cut search latency by 2.5 times, and enabled a more than threefold increase in ranking depth. The measurable business impact was substantial, including a 1.1% increase in transactions purchased and a 0.6% rise in Gross Merchandise Value, translating to over €3.5 million in additional merchandise sold. Vespa also halved the total cost of ownership, providing a cost-effective platform that directly boosted user engagement and revenue.

Vinted