Jan 16, 2013
I’m a nerd. I get on my bike or I go for a run, and I’ve got my Garmin GPS running the whole time telling me how far I went, how fast I went, what my heart rate was, (ok I’ll stop the list well before it gets to 20 items). I also happen to weigh myself 4 times a day. I like to know how much I weigh, what my % of water weight is, how fat I am, etc. I’m a total nerd and I like my data – lots of it.
When I get home from a routine Saturday ride, the first thing I do is download all the data. The data by itself is interesting, but only for about 3 minutes. The second thing I do is I trend the thing. I’ll look at my ride side by side with the prior week’s and maybe several others. Basically I’ll compare the stats and get an idea if I was slower, faster, more powerful, etc. In other words, within a few seconds, I’ll know if I’m better or worse off.
The thing is, I really don’t care about that either. What I really care about is that I was really slow up Mt. Tam, or that I got dropped on the way into Pt. Reyes. It’s not that I care that I suck (that’s a given), it’s that where I suck tells me what to work on. Sometimes, it even tells me that I didn’t eat enough (a common problem if you guys know me – yes, I’m neurotic).
In the last 15 years, most of us have gone from minimal data in our reports, to some pretty decent analytics and/or dashboards. We’ve started moving away from the static and columnar operational reports and into trending and drillable analytics. With these new reports, we’ve prided ourselves with the ability to tell our executives that “turnover is down compared to a year ago”, or “the cost of hiring in #BU has skyrocketed.” It’s a wonderful day in the neighborhood!
But I don’t want operational reports. Neither do I want analytics. What I want has almost nothing to do with the data. I want insightfulness into the business.
Let’s pretend I go to the head of my business unit and tell her that turnover is up over the last 3 quarters. “Crap,” she says. “What is the problem and what do we do about it?” Joe HR stammers and says, “looks like we have an engagement problem???” Trends and drills are really nice things. I know I’m lying, our managers really do want this stuff. But they only want it because we’ve starved them for data for decades. They really are only mildly interested in the data. Just as I only look at my bike ride speed out of curiosity, I know that the number alone tells me zilch. Just as knowing my average speed compared to last week tells me only if I was a little better or worse, ignores conditions on the road, and the environmental context. What I really want to know is how everything fits together to provide an analysis that give me the insight to act and make a decision. I want to know what to act on, when, and how to do it.
HR’s job in data is starting to transition yet again. We’re moving out of the business (I hope) of creating trends and drills, and moving into the business of context. So the turnover trend looked bad. Now the questions is how we create a regression model around our data to figure out what the primary actors are for that turnover trend. Maybe for once engagement only has a small contributing score to turnover. Maybe what we didn’t know was that the leader told their entire BU that nobody was getting a bonus this year. Perhaps the cause of that was actually some severe cost containment driven at the corporate level. How about some good analytics here to compare the cost of total turnover to the cost of those bonuses?
At the end of the day, we have much cooler data. I’ll give us that one (and the vendors especially). It’s time though to stop being producers of just the data alone. It’s not enough anymore. Perhaps when half of HR were “generalists” a decade ago this was ok. But we’re supposed to be partnered with the business now and we still can’t really diagnose what’s going on, let alone how to fix things. Once again, let’s stop being producers of data and become analysts of data. We need to start producing some insightfulness.