Mining networks for connection, again

Years ago, I consulted to a company called Visible Path (later acquired by Hoover’s in 2008), that had built a social network analysis capability, figuring out who in a larger enterprise might know someone is another company that someone in the sales side might want to meet. In the case of Visible Path that analysis tapped into the implicit social network latent in Exchange email servers. But in today’s world we are connected explicitly in social systems like Twitter and Linkedin, as well as the contacts in or email.


Visible Path diagram


The notion of mining networks to open connections with targeted people remains a holy grail for sales people and recruiters in the business world, and in the past few weeks I’ve seen two promising products that leverage this sort of information for social CRM sort of applications.


EnterpriseJungle is a new startup that asks ‘who will you discover?’ and helps professionals discover ‘peers’ that they should be connected to, but aren’t. These may be other people in the same company, or possible sales leads in other companies. EnterpriseJungle pulls this off by tapping into the data available in profiles of other corporate users, and from solutions like Linkedin.

EnterpriseJungle homescreen

EnterpriseJungle homescreen

In the home screen above, you see a Company Matches in the upper right, public matches (people outside the company) in the lower right, and a stream of updates to the left. Note that the tool supports messaging with other connected users, and is making selections based on affinity analysis.

Here’s a profile:

User profile

As you can see, EnterpriseJungle has instrumented the profiles to bring together salient information — perhaps from existing company profile sources — and makes it straightforward to send a message or connect via Linkedin.


Nimble is both similar and dissimilar to EnterpriseJungle. Similar, because it does provide a similar sort of network analysis to find a path to a target connection. Dissimilar,  because it can operate as a Twitter and Gmail appliance.

When I first looked at Nimble last year, it was an attempt to create a single location to pull in your Twitter and email, and to provid a single history of your messaging with contacts. A noble goal, but one that was very frustrating for me, because of slow performance and sometimes odd behavior.

However, the tool has been largely reengineered to fix those issues, and also extended in several key ways. For example, there is now a Chrome extension that allows Nimble to tightly integrate with the Gmail user experience, pulling up social network data on email contacts while the user is reading and writing emails.

Screenshot 2013-11-17 11.26.32

Here you can see the Nimble contact of the originator, and at the lower left Nimble has pulled her Twitter stream into context. You can see other controls, like a period of time for ‘staying in touch’, a repository for notes, and other social CRM capabilities.

Along with the capacity to act as a Twitter client, Nimble also pulls shared connections from social networks who can act as intermediaries and make introductions, conceivably.

Screenshot 2013-11-17 10.42.12



The Bottom Line

This isn’t an in-depth analysis of either of these solutions, but a connecting-of-the-dots based on demos. There is a continuing interest in solving the puzzle that Visible Path and others companies attempted in the days prior to well-established open social networks, and now the awareness that this sort of data is waiting to be mined makes the value proposition for doing so almost inescapable. My sense is that EnterpriseJungle is more oriented toward intra-company social connections while Nimble is more of a social sales appliance, but it’s clear there is a lot of overlap, as well.

These four ideas might make Yelp a lot more useful

Yelp has announced the winners of its inaugural Yelp Dataset Challenge, and the four entries it chose actually seem pretty useful. They run the gamut from a technique to highlight key words so users can read reviews faster to helping businesses predict whether they’ll see an uptick in activity on Yelp. Having read countless reviews giving restaurants low ratings even though the food was good, I think the entry that extracts subtopics (e.g., food, service, ambience) from restaurant reviews might be my favorite.

When he said ‘big data’ he meant ‘social data’, so ‘social analytics’ might be better

The buzziest term of 2013 is ‘big data’ which originally meant large amounts of formerly inaccessible data, like what people are doing on their cell phones and personal devices while watching TV. Increasingly, the fairly straight forward analysis of relatively modest levels of data is being called ‘big data’, like crunching employee data and comparing with survey results, as in this piece in the NY Times:

Steve Lohr, Big Data, Trying to Build Better Workers

Google, not surprisingly, is committed to applying data-driven decision-making to human resource management. For years, candidates were screened according to SAT scores and college grade-point averages, metrics favored by its founders. But numbers and grades alone did not prove to spell success at Google and are no longer used as important hiring criteria, says Prasad Setty, vice president for people analytics.

Since 2007, the company has conducted extensive surveys of its work force. Google has found that the most innovative workers — also the “happiest,” by its definition — are those who have a strong sense of mission about their work and who also feel that they have much personal autonomy. “Our people decisions are no less important than our product decisions,” Mr. Setty says. “And we’re trying to apply the same rigor to the people side as to the engineering side.”

I think these results are unsurprising. Personal autonomy, domain mastery, and the respect of those that you think highly of are the three considerations in work happiness, and clearly, these derive from doing a job well. The surprising thing is that Google’s earlier biases were believed in the first place.

My friend Rawn Shaw got similar heartburn with the piece, and suggested that ‘social analytics’ might be a better characterization, likening this analysis of employees to the traditional analysis that companies have been doing on their customers, for years:

The big picture people often miss out is that this in concept is no different than doing detailed analysis on customer behavior over their lifecycle. This is easier in some ways since you can more easily reach the data, for a longer period of time, and in greater detail that can be mapped to individuals (e.g., employees), than you typically can get from the limited transactions with an individual customer.

It is also more complicated because there tend more databases with employee and process information across more departments and business functions than with customer information. Plus, there are more issues in accessing this information while trying to maintain privacy of employees. It can be done on a practical level.

I think ‘process’ could be substituted with ‘work’ in the last paragraph, but otherwise I agree with Rawn, and also with his observation that companies have been willing to spend the time and money to analyze their customer data for insight. It is only now, when businesses want to get another tranche of productivity out of the workforce, and they have loads of social data streaming through social tools, only now, when it is becoming relatively cheap are they trying to mine that social exhaust for insights.