by Scott Pollak

When dealing with human resources, it’s not enough just to have the numbers. You also need to know how to act on them.

How many CPAs does it take to screw in a lightbulb? Is this a joke? No, it's a metric. A "standard of measurement" according to Webster's Dictionary. Yes, a metric. The benchmark for the lightbulb question is one. I don't have any normative or empirical data to support this assertion. But it should take one CPA to screw in a lightbulb.

Now here's the rub: Should it always take one CPA to screw in a light bulb? At every company, and in every division, should it always take one CPA to screw in that bulb? It is the benchmark after all.

You can imagine sitting in an HR strategy meeting, putting together your management scorecard, and a key executive says, well, "Are our lightbulb-changing CPAs sufficiently productive?" The question isn't too different from more-traditional human-capital queries like: How satisfied are our employees? Are we getting enough productivity from our staff? Are we spending too much money to find the right employees? Or, Is our return on our human capital investment competitive?

As with all measurements, the real concern shouldn't be the result itself. HR departments and human-capital decision-makers need to be as concerned about the context and causation of the metrics as they are about the results. More importantly, decision-makers need to ascertain how to take action on the basis of those results.

Let's say you do measure, company-wide, and by division, your corporate aptitude at screwing in lightbulbs. Knowing the benchmark, and your metric, you can begin to identify problems.

If you find that it takes your organization five CPAs to screw in the lightbulb-one to hold the bulb and four to spin his stool-then you have a problem. If it takes you three CPAs to screw in the bulb, one to do the work, another to make sure he feels adequately compensated for his efforts, and the third to administer his life events, you also have a problem. On the flip side, if it doesn't take any CPAs, because your purchasing department won't buy any more lightbulbs due to capital spending being curtailed, you still have a problem.

In these instances, knowing that the benchmark is one CPA is very helpful, allowing you to cut costs, or better yet, improve revenue production.

However, if you work for a robotics company and had cheap access to lightbulb-screwing robots, and installed these machines everywhere, then using a whole CPA to screw in that lightbulb might be far too much.

You might find that the ceilings in your call center division are quite low, but in your ladder construction division they are quite high, and as a result you may experience different and explainable results in each division. Or, quite likely, you could find that two years ago, due to the stool-spinning problem, it took your organization five CPAs to shed new light. This following year, you got this number down to three, having stopped with the stool but still requiring additional resources to provide emotional and administrative support to the remaining bulb-screwing CPA. Now this year, you might be quite happy to cut that support in half, still having two individuals involved in the process. You will be twice the benchmark of 1one, but you're 33% lower than last year.

Real Benchmarks

More traditional human capital metrics aren't much different. Determining the metric is fairly straightforward. Through elbow grease, a group of willing peers, and some wrangling back and forth, you could find out, as Saratoga did, that the Total Voluntary Separation Rate is 10.5%; that the Employee Revenue Factor is $264,429; that the Total Cost Per Hire is $2,482; or that the Return on Human Capital is 1.31.

Undoubtedly, your organization will differ from every benchmark listed above, sometimes higher, sometimes lower. Sometimes your difference from the benchmark will be great, and sometimes little. Getting the benchmark and your metric is just the first step.

The robotics company may determine that its revenue factor should be higher than the average due to its ability to leverage nonhuman assets. Another company may find that voluntary separations are a big concern in the call center but not in manufacturing. So assuming you are tracking the right metrics (a subject for a future column), the real question isn't "what's the benchmark?" It's "what's the right measurement goal for my organization?" Once you know that your company takes five CPAs to make the change, and that the benchmark is one, you have only taken the first step. Ultimately you need to decide if you have high ceilings or if you're spinning the stool.