<strong>Smart</strong> Solution Data Marketplaces Data Marketplaces Old Data for Sale Do you know what your data is worth? Many private companies and public sector organisations could be sitting on piles of money they could use for other purposes– but where can they go to buy, sell or exchange this ‘used’ data? Data exchanges are an exciting new frontier in the realm of Big Data. n By Stian Overdahl 78
Data exchanges, or data marketplaces, could potentially unlock new revenue streams for companies, as well as lowering the cost of engaging in data-intensive activities, such as training machine-learning algorithms. Aside from exchanges, white label solutions for large corporates and their suppliers are another promising development. Any company can estimate the value of its data by calculating the cost of collecting it, its value to an internal commercial project, and its quality and longevity (will it have value in five years’ time?). In the era of data exchanges, these figures take on a new cogency as companies sell data sets, give them to trusted partners or acquire new data sets. Establishing a value for data on a public marketplace could bring big benefits to companies. For example, a project generating a significant amount of data could defray some of its costs by selling it, while a project where collection costs are significant could be done more cheaply by buying-in data. Sheridan Johns, head of ecosystem at Ocean Protocol, a blockchain-based data exchange, says, “People understand that data is valuable, and people are building valuable companies using data, but what are people going to pay for data? Putting data on our marketplace, allowing data to be Data Provider Compute Provider of digital and emerging technology at professional services firm EY. Buyers of raw data often find that suppliers collect and aggregate data in different ways and formats. A company that wants to provide a weather service, for example, can run into difficulties as it must obtain data from various measuring sites. It’s likely that a weather station run by a professional organisation, like an airport, would collect data more frequently, whereas others, such as a community organisation, would sample less frequently. In addition, the various measurers may calibrate their sensors differently. Maintenance can also be a factor – for example, one supplier may experience more downtime or it may decide to upgrade its software, making the data fragmented or unreadable. There are also questions about whether a company can build a product with a competitive advantage by relying on third-party data sources, or does it need to build its own sensing infrastructure? “In 99 percent of cases, you will recognise that the competipriced in real time using dynamic pricing features and automatic market makers, enables an individual or organisation to publish a certain dataset and determine exactly what its value is.” Johns believes that within ten years it’s going to be commonplace for large organisations to be either buyers or sellers of data using online data marketplaces. This potential growth will result from the growing need for data, especially for artificial intelligence. Increased demand and a growing awareness of the importance of data also makes companies more likely to share externally, notes Didier Navez, senior vice president of strategy and alliances at Dawex, a data exchange. “In the context of training AI, it seems like the market is evolving and maturing, and companies are really willing to share data. Clearly data exchanges help in that process because they provide a structured environment where exchange of data is secure and compliant,” he says. There are still several flies in the ointment, especially in the domain of IoT: quality of data, privacy and commercial sensitivities. "When it comes to monetisation of IoT data, people think it should be easy. Unfortunately, there are a couple of huge issues and this concept very quickly becomes very complicated,” says Aleksander Poniewierski, global head Decentralized Data Market Protocol and Network of AI Data / Servies Storage Provider Analytics People understand that data is valuable, but what are they going to pay? Sheridan Johns Ocean Protocol Further Processing source ©: New Work SE source ©: Asia Crypto Today Algorithm f(x) Curation Building a Data Ecosystem Through the help of decentralised service agreements and access controls powered by virtual machines, Ocean allows users to access, monetise, and even curate data management services that they will need for their businesses. 79