Mine employee data to reduce turnover

Tools to perform HR analytics allow organizations to identify potential turnover problems before they occur

Employee turnover remains a costly concern at many organizations. Undesirable turnover imposes significant recruiting and training costs, plus the intangible costs associated with the loss of knowledge capital.

For these reasons, it is important to measure and predict turnover, understand contributing factors, and design programs for controlling and preventing it within targeted talent and knowledge levels. While this is not an easy task, the evolution of HR technology has produced new tools for HR professionals to improve retention.

HR analytics tools are an entirely new class of HR systems that aggregate not just HR but company-wide data.

HR analytics builds models that look for unusual patterns in data and statistically validate behaviours. While many HR systems can provide data about employee turnover, HR analytics provides insight into why the turnover is occurring.

Getting to the root of the problem

Understanding the root causes of turnover is not an easy task. It requires access to accurate, timely and comprehensive data about employees, managers, compensation and benefits, training, evaluation results and absence.

It is not about getting more data so much as analysing the right type of data the right way.

With the massive amounts of data existing in organizations it is likely that not all of it will be stored in one system. So, the first challenge is to access, collect, store and integrate data from all the appropriate data sources into one location for reporting and analysis.

Modern HR systems only operate and have access to the data they collect, manage, and store. And the HR system is just one of the many HR data sources.

What’s more, many HR systems have a significant data organization challenge. The data in these transactional systems is stored in a format to facilitate transactional and operational functions, and not business intelligence and decision support.

To continually make use of the gathered data, some organizations invest in an HR business intelligence solution which enables the organization to perform sophisticated HR analytics not possible with traditional HR systems.

From the data come predictions

Effective HR analytics enables the organization to analyze HR metrics and perform predictive modelling. For example, the HR team may want to compare turnover as it relates to voluntary separation, involuntary separation and churn.

They may then want to calculate the rate at which each type of turnover occurs based on employee demographics such as age, ethnicity, years of service, skill level or competencies.

Some HR systems can do this now, but will only highlight where the majority of turnover is happening and not provide any insight about the root causes. New HR analytics tools give users a thorough understanding of not only the trends that are emerging but why.

Anticipate and respond to changes

HR analytics solutions can deliver standard reports that measure turnover as well as portray relationships among selected employee characteristics and voluntary termination.

These systems can uncover areas in need of special attention, which the organization had not previously identified as a high priority area for HR retention and succession planning.

One HR department using a HR analytics solution discovered that a significant percentage of people with a critical skill was nearing retirement. The organization was not previously aware of this. These issues are especially prevalent in the public sector at the moment.

An HR analytics report shows the degree to which various characteristics such as salary, educational level, skills or length of service contribute to turnover. Additionally, employees can be individually ranked based on an assigned probability that they will voluntary terminate within a specific time window.

Applying advanced analytics, and specifically predictive modelling techniques, can rank employees to indicate the likelihood of someone leaving the organization.

Subjective factors like personality type and co-worker relationships have a bearing on turnover. There are also external factors that motivate employees to leave. These are all factors that predictive modelling can surface. In addition, there are ways to capture common reasons for leaving discussed in the exit interview — better salary, more room for advancement, a bad manager — can all be captured through coding and incorporated in a constantly evolving predictive model.

Once the behavioral characteristics of those employees most likely to leave are identified, it is possible to accurately anticipate changes and adopt plans to prevent them from leaving, or proactively mitigate the impact of those departures.

Succession planning can become more effective. If individuals identified as most likely to leave are part of the organization’s elite, strategies must be put into place to retain them.

Gary Love is a senior program manager with SAS Canada.

To read the full story, login below.

Not a subscriber?

Start your subscription today!