Tech Insights: Can Analytics (or Culture) Really Trump Politics?

Ten years ago, one retailer was far ahead of the curve on analytics.

It built a skilled analytics team the likes of which most of us would envy today. However in the end the team was disbanded.

The organization’s leaders felt their deep expertise in retail trumped anything analytics could tell them. Furthermore, they didn’t particularly appreciate it when some young analyst claimed the data didn’t support their decision.

When politics trump analytics

Sound familiar? It’s a normal case of politics (or to be kinder “culture”) trumping analytics.

We shouldn’t be surprised. When pioneers like Dr. David Sackett and Dr. Gordon Guyatt introduced the concept of evidence-based medicine, many doctors were enraged. The suggestion that doctors weren’t already being scientific or that their authoritative views on treatment should be trumped by clinical studies was not kindly received.

Organizations of any kind are political beasts and we can’t be successful in analytics without being successful in managing the politics. This can be hard because your typical “numbers junkie” is the last person you’d want to entrust with the task of navigating issues of power, status and culture.

Nevertheless, it has to be done.

4 ways to get leaders to accept evidence-based HR

Luckily, managing politics isn’t as hard it might appear. Here are four tips to get leaders to embrace evidence-based HR:

Involve — Managers want to be in control and they don’t want to be surprised. Involve them in analytics projects as much as possible and as early as possible. You don’t need to hand them the reins, just make sure they are involved and they know you are listening to them.
Support — Always position analytics as a way to support leaders, not as a way that usurps their authority. Find out what they want to accomplish and then provide analysis that helps them make good decisions. (Note support does not extend to cooking up numbers to create a smokescreen of support for bad ideas). When analytics support managers, managers will support analytics.
Frame positions as hypotheses — One of the best tricks I’ve learned from evidence-based guru Denise Rousseau is to frame positions as hypotheses instead of beliefs or opinions. If a leader says “I believe that people with a sports background are usually high performers” then you may have a problem because if the data disagrees then you have to go back and tell the leader they are wrong. You never want to make someone wrong. So instead replay the belief back to them as a hypothesis. “I see, you have an interesting hypothesis. You want me to investigate whether or not a sports background is predictive of above average performance. Great question. Let’s find out!” The hypothesis may be disproven, but the leader wasn’t wrong, they helped the company discover an important fact.
Recognize numbers are just one source of evidence — Sometimes data can lead you to exactly the wrong conclusion. So accept that not all data is high quality evidence and that if a manager has repeated experience in a field then that can be a reliable source of evidence. In the end leaders need to make judgements “all things considered”, so never treat data as if it represents infallible truth.

The good news is that data-based decision making has so much general acceptance that you are likely to get at least grudging backing for doing more analytics. Now, just follow those four tips and chances are you’ll be able to manage the politics that could get in the way.