Case Study: Catawiki achieves higher auction returns and faster ML deployment with Domino Data Lab

Maximizing Returns by Recommending Rare Products to the Right Audience

Catawiki is an online auction house for rare and unusual collectibles that relied heavily on a small team of expert auctioneers and a marketing/recommendation stack that couldn’t scale. Data scientists were constrained by limited compute resources and a deployment gap between Python/R models and a Ruby production environment, causing slow, offline batch predictions, missed personalization opportunities and delayed projects that risked losing customers.

By adopting the Domino data science platform, Catawiki empowered analysts to provision compute on demand, collaborate with stakeholders and deploy models to production without heavy engineering support. The result: auctioneers cut manual pricing research time by about 30% and improved price forecasts, marketing teams now target promotions and predict customer lifetime value more precisely to maximize commissions, models run as APIs with sub-200ms responses, and the company can run more auctions and capture greater revenue.

Catawiki

Peter Tegelaar

Chief Data Scientist