Written before about HR metricsand People Analytics(that’s a form of HR analytics), but for now I want to spend one second — well, one paragraph — on the overall idea of HR analytics. I’m a relatively smart person, but I’ve never led any HR studies or anything, so let me turn this one over to someone smarter than me. That would be John Boudreau of USC’s Business School, writing this post for Harvard Business Review about how People Analytics needs to be more user-friendly:
A good case in point is whether HR systems actually educate business leaders about the quality of their human capital decisions. We asked this question in the Lawler-Boudreau survey and consistently found that HR leaders rate this outcome of their HR and analytics systems lowest (about 2.5 on a 5-point scale). Yet higher ratings on this item are consistently associated with a stronger HR role in strategy, greater HR functional effectiveness, and higher organizational performance. Educating leaders about the quality of their human capital decisions emerges as one of the most potent improvement opportunities in every survey we have conducted over the past 10 years.
Houston, we have the nutshell problem.
Say it loud and say it proud…
Data is literally meaningless if it’s just collected. That just means you have data, which means people have more work to do trying to scrub/analyze it. Everyone is throwing themselves on the cross about their “added responsibilities.” Data scientists are being hired at 4.5x your salary. But it’s all trees falling in forests.
Data is only relevant if it affects decision-making in some way, and if it’s presented to executives in a way they can understand. This is why we need “data translators.”
Why is this seemingly rocket science?
Executives (“decision-makers”) at companies have been using specific vocabulary (buzzwords, acronyms) to describe what they do for decades. If the HR terms don’t match with those terms, they will probably care less — because their incentives and day-to-day schedules are tied to their terms, not vocabulary that HR uses.
Let me give you an example. If you teach an executive to think of talent sourcing as a supply chain, it will have greater business impact. The executive probably knows and can conceptualize a supply chain. He will “get” it. But if you go to him with lots of HR terminology, he likely will not care/dismiss it, because it’s not close enough to his “power core” of concepts.
Now, the easiest way around this is to have decision-makers who are adaptable and care to learn more about the business. Unfortunately, I wouldn’t call that normative. When you’ve worked at a place X-time and spent Y-time of that in one silo/division, you get pretty focused up on those terms.
Back to this HR analytics pull quote above
Here’s your main takeaway:
Whatever HR analytics or system you use, it needs to be tied to decision-makers having more info, easier to access info, and making better decisions.
Otherwise it’s basically being done in a vacuum. The HR analytics are nearly worthless.
Let’s use a common HR metric example here: employee turnover.
I realize very few companies use exit interviews, and the ones that do are usually pretty half-assed, so often the “data” you can glean here isn’t great. But here’s what you need:
- Costs of hiring/recruiting at different salary bands
- Costs of onboarding at (ditto)
- Turnover by department/quarter (and annual)
- Turnover by specific manager/quarter (and annual)
- Differences in these costs — i.e. what are specific managers and departments costing the company through turnover?
- Net promoter scores
- Any exit interview reasons you have gotten
None of this data is hiding anywhere. It should all be HR analytics that people can access relatively easily. Not brain surgery here.
So now what?
Now put together the data and what it shows:
- People leave at this clip…
- … it costs us this much money …
- … if 10% less people left, we’d save …
- … here are some other solutions/ideas …
Now you’re talking money. Executives probably will care more. You’re turning HR analytics into their language. It’s way more beneficial than trying to jam traditional HR methodology at them. Hate to break it to you, but most won’t care about HR speak.
Bringing it all together on HR analytics
What you need to do is:
- Collect data that appears relevant (you can test over time if it actually is)
- Organize/sort the data in ways that relate to cost/bottom line
- Present what you’ve learned in ways, and with vocabulary, that are amenable to the decision-makers
Lots of work, but only three big steps. Unfortunately I’m sure most HR/operations guys running HR will break this into 271 process steps, run everyone in circles, present a report that the executives all check their email throughout, and nothing will happen.
But it doesn’t have to be that way.
HR analytics can work. It just requires a little big picture thinking and some come-to-Jesus moments on data.