How to tell the difference between good and great HR analytics – part 1

Great HR Analytics

Question: What is the difference between good and bad analytics?

The answer is probably best illustrated like this;

This is bad analytics:
You have received the latest performance management data from all five divisions. You know from experience that the data is questionable – in two of the divisions many of the inputs are the default settings. You also hear that the managers have a somewhat relaxed attitude towards the accuracy of the data (to say the least). Your report show that absenteeism is flat at a reasonable level. You report this with satisfaction.

This is ok analytics:
You get the latest voluntary employee turnover data and see that the figures are trending upwards. The level is above the 6 month moving average. You dig deeper into the data and see that it is in particular in three division the employee turnover has been higher than expected. These division have three things in common; a new leader has been employed, workload has increased and they are client facing. You contact the relevant HR partner and suggest that they implement the usual retention initiatives.

This is great analytics:
You get the latest voluntary employee turnover data and see that the figures are trending upwards. The level is above the 6 month moving average and 2 %-point higher than sector-adjusted benchmark. You dig deeper into the data and see that it is in particular in three division the employee turnover has been higher than expected. These division have three things in common; a new leader has been employed, workload has increased and they are client facing. You then decide to interview relevant people in those divisions including leaders and employees as well as the HR partners. You read relevant academic research and case studies on effective measures against voluntary turnover in your particular sector and discuss this with experienced managers and HR partners within your company. You suggest three actions; coaching for the leaders, competency profiling to match job demands and team building. Your data suggest that these three initiatives will reduce the voluntary employee turnover from the current rate to 1 %-point below the benchmark at a ROI of 55%

 So what is the difference? Great analytics

 Any other suggestions to differences between bad and great analytics?

10 comments

  1. Great article Morten. Really goog!! But… are our organizations prepared for great analytics? Despite the incredible result that we can get, I’m afraid the effort needed is, so far, a quite a bit intense. But, who said no fear? Let’s try it!!
    Best.

    1. Hi Julián. Thanks for your comment.
      I believe that some organizations are prepared but most are not. They must be nutured and educated to see the benefits. I guess the only thing is to try.
      Best

  2. Hi, I miss one step… True analytics. Instead of reading academic research you create your own internal model for turnover. You have integrated the data of interviews, surveys, personality tests, recruitment, etc and build a predictive model that tells you what buttons to push in your company to reduce voluntary turnover.

    1. Hi Irma,
      Like your comment. Hadn’t really heard of ‘true analytics’ before but I suspect it is a predictive analytics model? Anyway, my argument for including other types of input such as experience and academic research is that I believe such input but complement a pure analytics model. I am unsure if we are there that predictive analytics models can stand on their own.
      Thanks for your comment – appreciate
      Best

  3. I would add great analytics “is actionable”. Would also note that generally the role of the analytics team will be to diagnose what is going wrong and identify what to change, but not how to change it as in your example.

    1. Hi Nigel,
      Thanks for commenting.
      Agree, great analytics must be actionable – good point.
      I would hesitate to separate the diagnostics and the what-to-change with how to change it too much. In my experience, there is a lot of valuable information in the diagnostics phase which is lost if those two are separated. Also, it would be useful for the ‘how to’-people to be able to run what-if scenarios with the analytics team. Analytics are many times ‘only’ doing diagnostics – I just don’t think that is the right way to do it.
      Best

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s