Big data is moving beyond buzzword toward ubiquity, and Accel Partners wants to capitalize on this trend by fueling the next generation of big data applications. With its new $100 million Big Data Fund, the VC firm’s offices across the globe will invest in applications that help form an ecosystem around existing big data building blocks such as Hadoop and NoSQL.
According to Accel’s Ping Li, exploding data volumes are creating an opportunity to form ecosystems similar to those that grew up around relational databases decades ago. Frameworks such as Hadoop and NoSQL become more popular because there’s just no other way to deal with petabytes of unstructured data, and new products are emerging that either build upon those frameworks or embed them into use-case-specific applications. It’s this next generation that Accel wants to fund.
For example, Li said investments such as those Accel continues to make in companies such as Cloudera and NoSQL player Couchbase aren’t part of the Big Data Fund. Rather, he’s looking for companies that add to Hadoop, NoSQL, solid-state drives or other building blocks by tuning them to become easily consumable analytics engines or by creating products to help users better visualize what their data looks like. Taking it a step further, Li is also interested in applications that might not play up their big data roots at all because the underlying technologies are just a means to an end.
Li noted Android-focused mobile security startup Lookout as an example, because although it bills itself as a security company, it’s really a big data company. It analyzes all sorts of data from mobile devices — 12 million of them — to determine threats. That job definitely requires a big data backend, he said. But what’s running below the surface is of no concern to end users, so Lookout doesn’t lead with what technology it uses.
It’s the same story I highlighted Monday in looking at how much money Hadoop is really worth when I noted that recent investments in Hadoop-focused startups don’t include the myriad companies that merely use Hadoop to accomplish their goals. I noted targeted-advertising specialists such as 33Across and security specialists such as ipTrust. Li referenced a CRM startup he’s talking to that’s using Hadoop as the backend for a recommendation engine to point employees toward the people they should be connecting with more often.
Anytime a company’s application relies on the collection and analysis of streams of data — structured or not — it’s a big data company at its core. Li thinks the mobile space will be a hot area for big data innovation because of the sheer number of devices — there are more phones than people in the United States — and the amount of data each device produces. The consumer Internet is another hot spot.
At some point, we’ll get to a place where Hadoop and other foundational big data elements are no longer novel, but ubiquitous, and “that’s the point of this fund,” said Li: to accelerate that transition. Just as we assume applications today are running Oracle, MySQL or some other relational database, we’ll just assume that certain applications are utilizing Hadoop for storage and processing, and it won’t be that interesting to find out.
In order to help out its new fund, Accel has roped in a variety of advisers, including bit.ly’s Hilary Mason, Cloudera’s Jeff Hamerbacher and Doug Cutting, and Facebook’s Jay Parikh. Additionally, Cloudera will be on hand to provide Big Data Fund companies with technical, market and customer support around Hadoop.