DaDaDa dataset: 16K data products benchmark pricing for marketplaces

Researchers from multiple institutions introduce DaDaDa, the first dataset specifically designed for data product pricing in data marketplaces. Traditional pricing approaches—cost, income, and sales comparison—fall short for data: cost fails because replication has near-zero marginal cost, income fails due to unpredictable revenue. Only sales comparison remains viable, but lacks standardized benchmarks. DaDaDa fills this gap by aggregating metadata for 16,147 data products from nine major marketplaces worldwide, including AWS Marketplace, Databricks, and Datarade.

The dataset enables training of machine learning models for pricing, classification, and retrieval of data products. Experiments demonstrate its effectiveness in establishing price benchmarks, and a retrieval prototype shows practical utility. By providing a standardized reference, DaDaDa helps marketplace operators, data sellers, and buyers arrive at fairer, more transparent pricing. The dataset and associated code are publicly available, encouraging further research and industry adoption. This work directly addresses a critical bottleneck in the growing data economy.