Are Big Data Startups Eyeing the Enterprise Already?

February 17, 2011 Off By David
Object Storage
Grazed from GigaOM.  Author: Derrick Harris.

Two popular big data startups made management changes this week, which might signal the companies’ boards feel they’re poised to make runs at the big time and need seasoned leadership to take them to the next level. At Karmasphere, which provides desktop tools to ease the process of writing Hadoop applications, former Omniture  and BEA Systems executive Gail Ennis has replaced Co-founder Martin Hall as CEO. At 10gen, proprietor of NoSQL database MongoDB, former MarkLogic and Oracle executive Max Schireson has come on board as president, reporting only to Co-founder and CEO Dwight Merriman, a respected IT veteran in his own right. These types of moves have proved successful in the recent past for similarly cutting-edge companies.

I’m thinking specifically of the cloud computing space, where BEA and EMC veteran Byron Sebastian came on at Heroku and former MySQL CEO Marten Mickos took the helm at Eucalyptus Systems. Heroku just sold to Salesforce.com for a staggering $212 million, and, despite speculationit will be buried by OpenStack, Eucalyptus has raised boatloads of capital and claims to be doing great (indeed, it did recently add offices overseas). The connection between all of these management changes is bringing in large-vendor experience to better position — and sell — high technology for business users.

Having a developer or scientist as the public face can be great for some audiences, but not usually for CIOs. They need to know the business case for adopting a new technology, not just the technical details, and they want business functionality, as well. The reality for startups in the Hadoop and NoSQL markets is that they’re at an inflection point, where large companies and large vendors are beginning to realize just how important these types of products can be for certain workloads. However, they want to see more than an awesome technology that requires a team of scientists to use and that only works for grand-scale problems. They want to see a product that the current IT staff can use today (or soon) for today’s problems, and they want to see proof others have deployed it successfully before.

This is no spectacular new insight, of course, just my thoughts upon seeing the management additions at Karmasphere and 10gen. They’re both companies at the forefront of their respective, albeit relatively young, spaces, and these types of changes can be significant. They suggest to me that Karmasphere and 10gen have established sufficient developer bases and are satisfied enough with the core technologies to start making their enterprise pushes. If they don’t succeed, they can’t blame management inexperience.

To get a better feel for how Hadoop and NoSQL are being used and what they’re capable of, attend our Structure Big Data conference March 23 in New York City, where both 10gen and Karmasphere will be on stage talking about getting a handle on unstructured data.