HOME > 人事コラム > Secrets of Human-Capital Management 【March 2017】

Secrets of Human-Capital Management 【March 2017】

How to derive insights from HR data

By David Creelman

 

In a previous article I discussed about how to get clean data for HR reporting. Now let’s move on to the next problem: deriving clear insights from the data.

 

Deriving insights

When I consult to companies on HR analytics, I find their biggest struggle is deriving insights from data. The reporting/analytics team provides reports, dashboards or self-service capabilities. People find this information interesting at first; but then interest fades because there is no actionable insight; there is no “So what?”

 

This failure may be the result of one of four factors:

 

  • 1.The HR reports may not be presenting the right data in the right way
  • 2.The users may not have the right skills
  • 3.There may be nothing of value in the data
  • 4.The whole process may be back to front

 

If you want to get better at deriving insights you need to understand which of these factors are at play in your organization and how to correct them.


Issue 1: Presenting the right data in the right way

When deciding what to put in HR reports we tend to ask managers “What data do you want?” This seems reasonable, but it doesn’t usually work because managers don’t know what they want. They’ll ask for a lot of things that sound vaguely interesting. This keeps the reporting/analytics team busy, however, in the end vaguely interesting data doesn’t get looked at. The reporting team’s hard work was for naught.

 

The key is to have a session with managers where you work them through a process of discovering “What information would you use? How would you use it?” Once you have a thorough understanding of how information will be used it becomes self-evident what to put in reports and how to display it.

 

Takeaway: Choosing what data goes into a report takes a serious investment of time up front. Fortunately, you win that invested time back because you won’t be creating reports no one uses.

 

Issue 2: Users may not have the right skills

The degree of “data savvy” of HR pros and other managers varies enormously. Certainly you should give your HR team training in the basics of “data judgement”. They don’t need to learn statistics; but they have to be comfortable engaging with numbers. Once they’ve engaged they will run into issues that are too hard for them; but that’s okay, they’ll recognize this and get help from the members of the team who have better training in quantitative methods.

 

Takeaway: Provide basic training for HR pros, but don’t try to turn them into statisticians.

 

Issue 3: There may be nothing of value in the data

To say that there may be nothing of value in the data borders on heresy; but it’s clearly true. Maybe you have terabytes of data on when your employees take breaks. That’s interesting, but there is no guarantee there is some deep insight to be found there. Yes, maybe you’ll find a certain pattern of breaks is connected to higher productivity; but chances are you won’t. Social scientists, who do this kind of thing for a living, know full well how hard it is to find significant new insights in big data.

 

Takeaway: Don’t invest a lot in randomly looking through data assuming something of value must be there. Unless you are pretty confident there is going to be value, then you’re probably wasting your time.

 

Issue 4: The whole process may be back to front

For years I’ve been training HR pros to start with the decision they are trying to make—and work back from there to the data they need; rather than start with the data (e.g. a report) and hope to find an insight that will help inform a decision.

 

Takeaway: Write up a clear process that starts with a business issue and an answerable question then works its way towards identifying relevant data that will help answer that question; and then give the HR team coaching in using that process.


Conclusion

There are various reasons why companies struggle with deriving insight from data. Perhaps the most fundamental reason is that we assume once we have some data, finding an insight will be easy. It’s not. We need to approach analytics is a systematic way with a clear understanding of what success looks like.


 

David Creelman is CEO of Creelman Research. He is best known for his workshops on People Analytics, Evidence-based Management and the Future of Work.  You can connect to Mr. Creelman on LinkedIn or email him at dcreelman@creelmanresearch.com

 

人事コラム一覧へ

 

 

みのりQ&A

人事用語集

みのり風土調査