An interesting reflection on this topic was published by Peter Cappelli, a professor of management at the prestigious Wharton School of the University of Pennsylvania. His article for the Harvard Business Review website was eloquently entitled "There's No Such Thing as Big Data in HR".
Small data, big problems
Cappelli explains that businesses have a maximum of thousands of employees and most of the data collected about them is valid for only one year. The use of special software to process big data is unnecessary. A bigger challenge for HR than big data is to learn how better to use data as such.
Data related to different HR areas, such as recruitment or performance management, is very often stored in different systems that are not compatible with one another. More than big data professionals, therefore, companies need database managers able to clean up and link their data. Common software tools, often even Excel alone, can do this.
Do not reinvent the wheel
Another important fact, according to Cappelli, is that the questions we are trying to answer using the analysis of big data were already answered long ago. As an example, he mentions Google's multi-year Project Oxygen research which aimed to discover what makes a good manager. Most companies could not afford such a vast data analysis. But what did it find? The same things academic research concluded decades ago and that we can read in textbooks.
Another issue concerns legal restrictions in the files of personal data protection. Multinationals are not allowed, for example, to analyse employee data across national borders.
What to do after cleansing databases of existing HR data?
Analyse relationships within the data. Start by what impact your recruitment criteria have on employee performance.
Remember, however, that data quality is the most important thing. If, for example, you are using a performance measurement system that you think is not a good measure of actual performance, it is pointless to analyse the data you collect.
What do you think about the use of data, or big data, in HR?
You can also read an interesting discussion under Cappelli's article here.
-kk-