Improving B2B Data Markets

Abstract

Markets for information are growing rapidly, thanks to the growing number of data sources. Many businesses have taken notice to the indisputable benefits of data sharing and are taking action to make their data available to potential buyers. Data is a non-rival good, and thus it can in theory be sold to every potential buyer. In practice, this isn’t the case. Businesses that wish to practice data sharing face many risks and obstacles. Data markets are of monopolistic nature, which allows every seller to greatly influence their pricing policy, instead of taking market price as given. A buyer looking to buy data from one or multiple sources faces potentially formidable costs. Sellers on their part could make the market more attractive by combining their data, but this requires a high level of trust between the sellers. These are only a few examples that prevent the market from reaching a socially optimal outcome. The European Union suggests using neutral data intermediaries, that could collect data from different suppliers, creating an effective single market for data. Contributing suppliers would be compensated with increased access to other’s data, analytic services or licensing fees. This solution constructs effective infrastructure for data markets but fails to set incentives which would induce desired outcomes. Suppliers have an incentive to provide low-quality data in hopes of gaining access to high-value data contributed by others. Additionally, the intermediary cannot price the data effectively without creating some sort of scarcity. I propose a model that utilizes the data intermediary solution and the VCG-mechanism. By collecting and combining the contributed data into a single homogeneous dataset, the intermediary can auction it by using the uniform price multi-unit Vickrey auction. Auctioning ݇ units of access rights to ݊ buyers, when ݇ < ݊, creates artificial scarcity which induces truthful bidding. The intermediary collects data from the sellers individually, not allowing the sellers to gain access to each other’s data, which reduces incentive to provide low-quality data. The intermediary will then compensate the suppliers with the generated auction revenues.