There are lots of words and concepts floating around the HR world these days that are new or different from the traditional approach – ones like “big data,” “HR analytics,” and “predictive analytics.” We think it is worth defining and discussing each of these topics.
Definition: A collection of data sets so large and complex that it becomes difficult to process using traditional data-processing applications.
HR Application: HR has tons of data in various systems (e.g., HRIS, ATS, LMS, surveys, 360 assessments, performance management ratings, compensation data) but we don’t always harvest the insights and intelligence from the data. Even though our data sets don’t always have the magnitude of data as traditional big data applications, the big data principles still apply!
Example: A good example of big data outside of HR is in consumer banking. Consider the number of individual customer transactions a bank processes every minute or hour. The big data concept started outside of HR, and most IT professionals would NOT consider the typical HR analytics project and data sets as “true” big data. (Although, at SMD, we have analyzed many data sets with millions of data points.)
Connection to Analytics: Underlying all the big data talk is the concept of analytics. Analytics provide the methods to harvest intelligence from the data. So, the HR analytics concept is much more important for HR professionals.
Definition: The analysis and application of a company’s people data to uncover people insights/intelligence that informs HR strategies, process changes, and investments – all with the goal of improving organizational performance (i.e., driving actual business outcomes).
Deep Dive: HR analytics come in many forms – from basic descriptive analysis and tracking trends to the more sophisticated predictive analyses like multiple regression and structural equations modeling. This does get rather technical but understanding the basics will help all HR professionals be better consumers of analytics. The model below outlines four levels of HR analytics. Organizations should be utilizing and operating at all four levels. However, it should be noted that the most impactful analytics fall into the highest level – predictive analytics. Get an even deeper dive in our recorded webinar, “HR Analytics Step-by-Step Series #1 – Foundations.”
HR Application: HR has become much more sophisticated regarding metrics and data collection. However, most of our metrics are focused on HR efficiency (e.g., time to fill) and comparisons to benchmarks (e.g., engagement scores). The next evolution for HR is to directly connect people data to business performance via HR analytics.
Example: For example, we can identify the root causes of turnover in the first six months. The analytics can be used to understand potential impact and actual ROI over time. Ultimately, the analytics provide the business case, path to reduction, and the “story” of impact and value HR brings to the organization. We used turnover as an easy example, but we can do this with any business metric (e.g., sales, revenue, safety incidents, customer satisfaction). Bottom line – we must harvest the power of analytics to demonstrate the value of HR by driving organizational performance.
Definition: The application of analytics to existing data sets to predict future outcomes and trends.
Deeper Dive: Predictive analytics don’t need to be applied across the board in HR. This type of analytics requires a higher level of resource and investment. However, predictive analytics are the best approach for solving complex business problems. The analytics method does matter in predictive analytics. Neither data visualization nor correlational analysis are predictive.
Buyer Beware: If a vendor tells you they use predictive analytics based on correlations, they are wrong! Correlation means two variables are related but makes no inference regarding the order things occur in (a basic requirement for predictions) nor the causal implications of the relationship. Predictive analytics must be based on some form of regression analysis or structural equations modeling (SEM). In fact, one can infer cause and effect from an appropriately-run SEM. Okay; enough of the nerdy stuff ….
HR Application: In HR, we are generally looking to use predictive analytics to understand the people causes or drivers of meaningful organizational outcomes.
Example: For example, what employee experiences drive voluntary turnover or what competencies drive quota attainment for sales professionals? Based on the analytics, we can uncover the people levers to reduce turnover or drive quota attainment.
Hopefully, these basic definitions and concepts will help you become a better consumer of HR analytics, which is important because a lot of VC money is being funneled into the HR analytics space and some of the new products are being oversold or even worse, are garbage. Ask the hard questions and remember it is good to be skeptical in a time of great change. We have some thoughts on what will happen in the future, listen to our webinar here. Even MORE insight is available in this interview conducted by the Analytics in HR Academy with our Director of Research & Analytics.