1 Comment

Summary:

A few providers are not just creating huge data dumps — collecting and structuring data from multiple sources — but creating marketplaces for others to find and use that data. They include Microsoft’s Windows Azure DataMarket and InfoChimp, who hope to connect enterprises with the data they want

Moe Khosravy of Microsoft and Flip Kromer of Infochimps at Structure Big Data 2011
photo: Pinar Ozger

Moe Khosravy of Microsoft and Flip Kromer of Infochimps at Structure Big Data 2011Big data sounds good in theory, but where’s the money? A few providers are not just creating huge data dumps — collecting and structuring data from multiple sources — but actually creating marketplaces for others to find and use that data. Whether it’s a big player like Microsoft and its Windows Azure DataMarket or a startup like InfoChimps, there’s a great opportunity to provide a way for enterprises to be able to access vast numbers of content stores.

But how big of an opportunity? When asked at Structure Big Data how big data marketplaces could become, InfoChimps CTO Flip Kromer quipped, “We’re the next Google.”

Moe Khosravy, the product unit manager for Windows Azure DataMarket, was a little more measured in his response. “We wouldn’t enter the business unless we saw it as a significant opportunity,” Khosravy said.

Those companies are taking a slightly different approach to the marketplace model: Kromer said InfoChimps looks at itself as the “Amazon of data, “creating an open marketplace where anyone can share his data for free, or charge for it. Meanwhile, Khosravy said Microsoft is focused on high-quality data and has set its terms to ensure that.

Either way, there is clearly a demand for what they’re offering. According to Khosravy, there’s currently friction for app makers and others who want to use data from multiple sources, but are unable to do so through a single consistent API.

There could be even more of an opportunity as certain vertical data industries become more mature and there needs to be someone in the middle to enable enterprises to sort through that data. Khosravy noted that right now you almost need a marketplace for marketplaces, since different verticals have different APIs and ways for how their data is stored.

“What we need is a generic platform for data across industries,” Kromer said. “Soon every industry will come online as a mature data industry.”

Watch live streaming video from gigaombigdata at livestream.com

You’re subscribed! If you like, you can update your settings

  1. I prefer “One person’s metadata is another person’s data” for a headline ;) I was on a panel with Moe and Infochimps at the O’Reilly Strata Conf

    The notion of a data marketplace is a bit of an unfounded myth and the analogies to an App Store or Amazon don’t hold water–marketplace means buyers and sellers with a commoditized or well understood products: equities, pork bellies, gold. With differentiated products there is little transparency between buyer and seller. Purchasing books on Amazon are well understood. Music is in a common format. Both have millions of buyers.

    There is great heterogeneity across data sets (a term that is also not well-defined), so comparison/evaluation becomes a challenge on its own. Hence the value in curated offerings. ‘Demographics’ alone can be 2,000+ indicators.

    A data catalog might be more on-point, but is this an FTP server in the cloud?

Comments have been disabled for this post