Analytics Can Help You Optimize Your Comp Plan In Ways No Survey Can

Article main image
Aug 9, 2019

No field in HR is more numbers driven than compensation, yet it seems to be an area the people analytics team usually overlooks. There is an opportunity here for those companies willing to seize it.

Much of the number crunching the compensation department does isn’t what we’d normally call analytics; it’s more like budgeting or finance where the goal is to be sure the numbers are calculated and reported correctly. There is analysis in activities such as comparing internal salaries to the market, as well as analysis of how well actual performance pay matches policy. That is important work, but there is much more significant work to be done.

What questions should the compensation team be tackling?

Two big questions compensation departments should be answering are:

  • What would happen if we paid this group of employees 10% more?
  • What would happen if we paid them 10% less?

Changing the pay level of a group of employees relative to the market can have multiple large effects. It will have an impact on the quality of employees you can hire, the speed with which you can hire then, and how quickly they turn over. Ideally, companies would optimize pay against these outcomes for different groups of employees (i.e. you might have a different compensation tactic for warehouse workers and clerks). What I gather from my research is that it’s exceedingly rare for compensation departments to do this kind of optimization. More typically, companies simply decide what they will pay relative to the market average and leave it at that.

Another important question is: How much more should we pay our best performers relative to others in the same job?

For some jobs, like sales, we do see significant pay differences based on performance, but classically the top of a pay range is only about 50% higher than the bottom (e.g. bottom of range = $40k, mid-point = $50K, top of range = $60k). If you look at a championship sports team, in this case the Stanley Cup winning St. Louis Blues, they pay their top right winger 10x what they pay their bottom one (i.e. their best forward is 10x more valuable than their 12th best forward).

For jobs in hockey, where performance is easy to measure, it makes sense to pay far more for your top players. Are there jobs in your organization that follow the same dynamics? Chances are no one really knows. If a 10x (1000%) differential is ideal and you have a 50% differential, then it’s a terrible compensation policy.

Dealing with the difficulty of the analysis

Answering the really important questions in compensation is difficult. It’s easy to say, “Let’s optimize pay policy levels,” however there is a lot of research and experimentation needed to get even rough estimates on how it will impact factors like quality of hire (not to mention the need to put a dollar value on that extra quality).

However, organizations are already making optimization decisions based on hunches such as, “Let’s just pay at the market median.” That’s not a terrible strategy, however one would imagine that a good analytics team could do much better.

With so much money at stake, any improvement will easily pay for the investment in analytics. We should be tackling the high impact areas of people analytics, not just the ones that are easiest to work on.

Image by Gerd Altmann from Pixabay