While many people have an unwavering reliance on interviewing and reading past records and evaluations, few HR professionals would argue that a data-driven approach is now indispensable in their work, because all other approaches - personal and impression-based - offer limited insight into employee relations and can easily become subjective.

KPMG's own data-driven HR methodology offers a solution to these very problems. When creating the system, we aimed to be able to divide large groups of employees into groups from an HR perspective, to be able to reveal to management what type of employee groups (person zones) they actually work with.

Although most managers think they know their employees well, the results of our methodology have surprised many senior and middle managers. In fact, human characteristics and motivations are not easy to see through, and it is even harder to find patterns in them that influence their functioning positively or negatively, and in general, we are different from what we see ourselves as. This is particularly true when it comes to mergers of two companies, mergers of large departments or other restructuring, because employees who are feeling insecure, at least on the surface, show a different face because of the stress or perceived expectations of their managers.

The method consists of a questionnaire covering several company values, operational characteristics and behavioural dimensions, which can map the HR-determining characteristics and preferences of employees without the respondents being able to identify desirable or undesirable answers and therefore without being able to consciously influence the outcome. By analysing the resulting database, we can then answer questions such as what values, leadership behaviours or work habits are important to employees, and which will help to improve employee engagement and performance.

From the data - and here we are talking about questionnaires from hundreds or even thousands of employees - the analysis reveals clusters of people with similar characteristics and attitudes that would not be visible to the naked eye, as these nodes are visible on a map with many dimensions and hundreds of variables.

What these groups ultimately have as defining characteristics is jointly identified and analyzed by KPMG's data and HR methodology team. Through iterative reconciliation, each employee group becomes a "person zone" at the point where correlations between data points can be kernelled into clear trait and demographic groups, so that a description of common person zone characteristics helps answer key questions for employers.

The common characteristics of the “person zones” are not simply demographic data, but rather a set of common traits such as motivational factors, control tolerance or control-seeking tendencies, extroversion-introversion, creativity, or the desire or fear of change. These qualities are crucial for any change management task, and the reinforcement and communication of common strengths and preferences, or the reduction of factors and fears that make change difficult, are an important part of successful change management.

Our methodology relies on machine learning methods, as managing a multivariate space without recognizing patterns, subjective biases and creating groups would be impossible without it. And in terms of change management development directions, we are already working on incorporating elements of the neuropsychological knowledge domain into our methodological approach, because we believe that only methodologies updated with new empirical knowledge can create added value for our clients and our own colleagues in the long run.