HR metrics – each selection has consequences, good and bad

HR metrics have been the topic of last week following a Bloomberg article which has been widely discussed. However, for me the clearest voice has been Morten Kamp Andersen of Aspector who wrote a great blog post called ‘Why HR KPIs still matter but why they still fail to deliver’. So on Friday I picked up the phone to have a chat to him about metrics / KPIs, why HR is where we are and what we do to improve.

Morten spent the early part of his career in London as a Financial Analysts for the likes of Merrill Lynch and Deutsche Bank. Now back in Denmark he advises predominately Danish clients on Human Capital strategies including metrics / analytics.

I started by putting to him that KPIs are powerful if done well, but too often are not well executed.

“They are everywhere, people use too many and they’re simply bad. What happens is people meet often just once a month to see how they’re doing and the rest of the time they go unmonitored. 

But that’s not the main problem. Doing a KPI is something that can be dangerous if not done well. People naturally want to follow KPIs and there’s always a tradeoff with every KPI. 

Let’s take recruitment as an example. If time-to-fill is one of your KPIs and you’re going to be discussing this at your annual review then say you’re faced with 2 candidates; the first a the better option but with a 3 month notice and the second not so strong but with a 1 month notice, you have an incentive to take the inferior one but it’s not in the best interest of the company.”

Of course the other KPI which is often used is cost-per-hire and here there’s a trade-off with service. If you’re hiring for a banking front-office then you probably want to increase the service as your population is likely to be very valuable. Of course the easiest way of reducing cost-per-hire is to focus on the denominator and increase the number of hires as you have fixed and variable costs.

“Yes, and you also get pressures to focus on less-great channels.

My main point is that for every KPI you choose there will be an upside and a downside. There will be something you miss by the choices you make. This shouldn’t put you off making selections, just that you should understand the consequences in terms of behaviours of the choices you make. 

So for example your recruitment expert might pick 3 or 4 KPIs but they should understand what are the consequences of this choice and are we happy with that?

I don’t believe that many people go through this process of understanding and rigorous selection”

This brought us back to where I had started with this set of articles, using metrics in combination to describe a frontier curve or space to illustrate the possible choices at a certain level of efficiency.

“There is no doubt you should only have a few. If you have a too long to-do list the danger is you will get satisfied with ticking off items and end up picking the ones that are easy to do or you like to do – one tick is as good as other.

If you have too many KPIs the same happens. You focus on the ones you prefer not the ones that are most powerful. So make sure you select only a few.

The second is make sure that the KPIs complement each other. Even then there some aspects and things that fall through. That’s just how it is. You need to understand these implications up-front.”

So why do people pick the wrong KPIs? Morten reckoned that KPIs should be the last thing that are chosen when you make a change. When you understand what you’re trying to achieve the KPIs should be obvious.

The problem is that they’re often chosen at the beginning of a project in the business case phase of a project and ‘these things stick’.

My own view is that this is partly to do with the less rigorous use of decision milestones in projects: that people are given targets (or should that be KPIs) on delivering a project in their review and if further analysis shows that the original assumptions weren’t correct there is an incentive not to change. I guess this is yet another example of the incentivising behaviour of KPIs.

Finally I wanted to ask about Human Capital investor metrics, given his previous experience as an analyst. I’ve recently been advising a prominent sustainable investment firm on the Human Capital element of their methodology and thus the questions they ask firms. This week I have questioned whether they offer the panacea that has sometimes been given to them.

“Certainly simple metrics have a big pay-off. Just look at P/E. It’s hardly the most resilient metric but everybody uses it and interprets it depending on their own investment approach. I think we’ll see the same with human capital metrics. They’re a tiny step towards making HR more relevant.”

Certainly time will tell, but let’s ensure we understand the behaviour changing conseqents of our metric selection.