“We can’t do that, it’s personal information!” That was the response I got from a senior executive at a large corporation when I asked whether they were using social data from their employees as part of an HR analytics project.
This executive’s group was studying headcount, attrition, and employee engagement and attempting to correlate these metrics with overall business performance. They were utilizing two primary data sources – PeopleSoft and a quarterly employee survey – and trying to predict things like voluntary turnover based on responses to the survey. I suggested using additional data sources, like Facebook and Twitter, in an attempt to gather more information about employees’ intentions. He reacted as though I were suggesting he pick the lock on his teenage daughter’s diary.
While snooping in a diary may be a breach of privacy (at least from the teenager’s perspective), monitoring people’s Tweets and Facebook posts may not – at least according to several of this week’s articles, including our pick of the week from the Wall Street Journal.
This ongoing debate about privacy is an important one. On one hand, users have a right to know if – and how – their personal information will be used. On the other, app makers know they are sitting on a proverbial goldmine of user-provided information that can be used not only for their app but as a revenue source when packaged and sold to interested third-parties.
This information can be useful even if aggregated or made anonymous. But it gets even more useful if the data can be tied to specific groups of individuals or even to the person-level. Is this “crossing the line” as the HR executive would suggest? Or is it “being smarter” about your data?
This executive showed us an interesting anomaly in his current analysis: Employees who left voluntarily actually scored higher on the survey question “How happy are you in your current role?” than those who remained. So people who said “I’m really happy” ended up quitting within 90 days, while those who basically said “I’m miserable” continued to stick around.
There are some potential reasons for this; maybe the happy people are better performers and were able to find new jobs, or maybe people weren’t answering the question truthfully. Understanding the underlying reason is important, especially if you’re going to have HR staff spend time with employees who are in danger of leaving – do you have your staff spend time with people who scored low or high?
And understanding the real reasons behind the behavior requires additional data. One potential source of this data is social; if people are posting and tweeting “I hate working for Company XYZ” it may help understand whether your survey data is valid.
By itself, social data is not the sole answer. There’s a huge amount of it, it’s difficult to separate the wheat from the chaff, and it’s hard to get information at the right level. But combined with other sources, it can be a powerful addition to your analytics environment. Keep that in mind the next time you send a tweet mentioning your company’s name.