Real Analysts Don’t Cherry-Pick the Data to Make Your Case

Article main image
Apr 4, 2019

One of the most common requests HR leaders make of the analytics team is that they want to make a case for some initiative with data. It’s inevitable that HR leaders will want to make a case for initiatives and it’s a good thing they want to use data; the trouble is that it can undermine the core purpose of the people analytics function.

The point of the function is to lead to better decisions. The decision could be whether a given intervention, such as a recognition program, is worth doing. Or it could be to choose between several alternative courses of action. Analytics can help with almost any conceivable issue HR needs to make a decision on and that’s a wonderful thing.

However, as soon as the mission is to provide some numbers that will justify a program, then it’s unlikely anyone is really doing analytics at all.

The point of analytics is to dispassionately look at the facts and decide based on those facts. If the mission is to justify something, then the process is hardly dispassionate; the decision on what HR wants to do has already been made.

Organizations counter cherry-picked facts by putting HR through the wringer; questioning their numbers, questioning their logic, and questioning their conclusions. This kind of critical inquiry is useful; however, we’d be better off if instead of setting up an adversarial system, both sides were simply interested in using data to make the best possible decision.

Let’s imagine HR is keen to run a new team building exercise but the people analytics team show this won’t be effective. HR should be delighted! They shouldn’t be pressuring the analytics team to go out and find numbers that will better justify the program. They should be eager to trash that idea and work with the analytics folks to find a more effective approach.

If you live in an organization that is deep in the “make a decision, then cherry pick data to justify it” mode, you face a difficult process of culture change. We also need to admit that we’ll never be entirely free of this tendency. Let’s just remember that if you are justifying a decision that’s already been made then you are not doing analytics. Let’s get to a level of maturity where cherry-picking data is the exception, not the norm.

As always, thanks to our community of practice for their ongoing insights into how to effectively apply people analytics in organizations.