For the first year of Mailana's life, I was focused on building a system that large companies could use to identify internal experts, based on the content of email messages that their employees sent. What I built was effectively an auto-generated LinkedIn, and you can see a demo of it here. There was a lot of interest in the idea, but it floundered on both privacy concerns and the fact that only a small percentage of people in a company spend time looking outside their immediate team. I had plenty of mechanisms to ensure people's information stayed under their control, but it still felt a bit creepy, and definitely freaked out legal and compliance folks. The real killer was that the people who drooled over it, the internal entrepreneurs and the uber-salesmen, were not the people writing the checks.
After I switched to applying the same technology to the consumer world, I still kept an eye on my competitors' progress. I've been pretty sad to see a lot of ERMs flounder, though there are still a few fighting on:
Tacit was a big, late 90's traditional enterprise software company that received a lot of investment. They were building on a very similar concept to mine, finding experts on technical topics based on mail messages. I heard some good reports from their users, but was also told their interface for contacting experts was extremely clunky, apparently because they had a lot of mechanisms to preserve privacy that got in the way of the user experience. The company closed and the technology was sold in a fire sale to Oracle a couple of years ago.
Another well-funded startup, VP spent six years in the '00's focused more on the connection side of corporate social networks, trying to offer value through identifying strong relationships both inside and outside the company. It was an interesting contrast with LinkedIn's approach of capturing any connections you have, with no way to differentiate between childhood friends and people you met once at a trade show. Unfortunately they hit the same sort of issues I did when trying to sell enterprise-wide systems, and were moving to a more individually-based product when they were bought out by Hoovers. I was very sad to see that the product has now been discontinued, it had some rabid fans.
Bought out by Thomson/Reuters a couple of years ago, Contact Networks was also focused on mining the contact information that lives within an organization, but without worrying too much about the strengths of any connections. They still appear to be going strong with some recent updates on their site.
A firm I discovered when they attended the first Defrag, Trampoline were looking both at identifying experts and internal relationships. They used to have a great demonstration of their Sonar platform using the Enron email data set, but unfortunately that seems to have been taken down. They're still pushing ahead with their work, and recently have been looking to an innovative way of raising money that they've dubbed crowdfunding.
This was an innovative experiment by MS a few years back, rolling a lot of these expertise and relationship mining ideas into a prototype Outlook plugin. The add-in was removed after a few months, but I hear some of the same technology is finally making it into the latest versions of Sharepoint and Outlook.
So what does the future hold? I think Microsoft's moves are a good indicator. We've now got a world where more and more social network features are being accepted within the enterprise, and internal services like Sharepoint or Jive are the natural distribution channels for this sort of work. There's always been people who love having this sort of information to help them in their jobs, the problem's been getting it to them and then getting revenue back in return!
Pete Warden's blog 