For many years now, recruiting, talent management, and HR professionals have been using metrics. However, more often than not, the story these metrics tell is irrelevant or disappointing. One of the biggest issues is that almost all HR metrics are historical, and executives are currently more concerned with metrics that are forward-looking and can alert managers about upcoming opportunities and threats. In order to really gather decision-enabling insight from metrics, organizations require intelligence that looks at the entire time horizon, so that they can learn from the past and forecast for the future.
While many executives are focused on forward-looking metrics and predictive analysis, it is important not to disregard the value of reporting on historical data. The use of audit metrics is important to measure the effectiveness of HR programs and activities. Efficiency data can provide insight on key topics such as how well an employee is performing or how long it takes to fill a vacancy. However, HR organizations cannot rely on these metrics alone, and need to move away from their tendency to focus only on historical data, in favor of forward-looking analytics.
One of the key challenges talent organizations face is making the right choices about the conversion, usage, and storage of historical HR data. Companies have volumes of historical data about their workforces, and often struggle with the decisions about how much and what type of data to use. Having too little data can hinder the ability to generate the reports and workforce analysis needed, while too much can have an impact on the cost, effort, and length of data analysis projects. Interestingly, it is important for companies to consider the future when looking at their historical data. Understanding the organization’s specific requirements and actual needs for historical data, both now and the future, helps to guide which data to use and which historical metrics are relevant and useful.
And it’s important to keep in mind that historical reporting cannot solely enable decision-making, whereas historical data is absolutely necessary to conduct effective predictive analysis. Popular historical metrics such as turnover rate, benefit cost per employee, and cost per hire can be used to design workforce analytics that will help HR departments improve their understanding of worker demands and desires and focus on the future.
“If you can collect a lot of data about the workforce and look at it holistically, you can predict who the right people to hire are, and who are most likely to be successful as leaders. If out of the last 20 people we hired in this job, the four people with this background failed, we're not going to hire people with that background again.” ~Josh Bersin, Founder of Bersin by Deloitte
Forward-Looking Metrics Rely on Historical Data
According to Bersin by Deloitte, to truly gain value from predictive analysis and workforce analytics, HR managers need to have a solid understanding of business goals for the short-term and the long-range. This will help them collect pertinent historical data, apply relevant parameters, and extract insights that will help the company make decisions.
When building a predictive analytics program, it is important to start with the problem rather than the data. Understanding what big decisions need to be made and what problems need to be solved allow organizations to sort through their volumes of data to pinpoint exactly what they need to answer these questions.
The best forward-looking analysis incorporates historical reporting by telling managers why things happened, when they happened, and what can be done to overcome them. Many HR organizations are still laying the foundation for the future by standardizing data and definitions, improving historical reporting, automating metrics, and adopting new tools.
“When an organization is ready for predictive modeling it will need enough history to build the most sustainable model.” ~Rajan Dutta, Director of the Metrics and Predictive Analytics Practice at PwC Saratoga
Examples of Forward Looking Metrics and the Historical Data They Rely On