The time and attendance system in a 10,000 person company captures millions of transactions each year.
That data drives core processes such as payroll. But can the millions of data points do more? Can we dive in and get fresh insights about how to run our operations?
I asked Gregg Gordon, Senior Director Big Data Practice, and Iva Zafirova, Principal Strategic Consultant at Kronos, what they’ve found about extracting actionable information from workforce data.
A window into managerial behavior
They use a tool called Workforce Auditor to find weak signals in the data; signals that cannot easily be picked up by, for example, by just sorting on a factor like “time card edits” from high to low. There are 16 different types of edits to time cards (that number surprised me) and looking at multiple factors using a specialized analytical tool may pick up patterns you otherwise wouldn’t see.
When Gordon throws out a term like “K-means clustering,” you know they are doing something you probably wouldn’t do on your own with Excel.
They found some interesting things. For example one client’s data showed a store manager who was frequently deviating from the recommended schedule — and getting above average results. She knew something that the scheduling program didn’t. At the least, this insight will allow the company to tweak the scheduling software so that she doesn’t have to waste time changing the schedules. On a deeper level she may have insights about efficient scheduling that are relevant to many stores.
The same kind of analysis can also pick up managers who are deviating from best practice, gaming the system or breaking compliance rules.
For example one study found a manager who was regularly amending time cards just enough to avoid paying employees earned overtime. The company could now step in and have a word about that behavior. On a more positive note, the analysis has also identified managers whose patterns of behavior were well-intentioned, but the data indicated they would benefit from coaching.
We’ve come a long way from the days when time and attendance was the least interesting process in HR. It is now, if we choose to use it, a valuable window into managerial behavior and operational effectiveness.
What is interesting?
- With time/attendance and scheduling software (i.e. workforce management) we have masses of data, but knowing what to do with that data isn’t always obvious.
What is really important?
- The vendors are doing the really hard work for us; they are the ones with the data scientists figuring out how to extract meaning, and the programmers designing friendly interfaces so we can get at that meaning. You could wait until everyone else is using these analytical tools, but surely life will be more fun if you go out and give them at try.