How people analytics and AI are changing HR departments

Using technology to permeate into employee activity and behaviour brings new ethical responsibilities, say Dr Neil McBride, Dr Mayen Cunden and Vincent Bryce

Human resource departments are changing. What was an administrative function, running payroll and managing recruitment and training, is now a key player in corporate strategies and a major influence in employee’s everyday work experience. How has this come about? The answer lies in people analytics.

People analytics (PA) applies the power of artificial intelligence (AI) to the large data sets about people held by human resources in order to solve business problems. If a company has a problem retaining good staff, PA will tell them why and what to do about it. If sales in some shops are not as good as others, PA will identify the root of the problem in staff engagement so that the company can change managerial behaviour.

The range of data available to PA includes not just HR staff records, but an increasingly wide range of data types which could have movement and health data from wearable devices. Using machine learning, HR departments can identify new trends and influence people management decisions across the company. 

The pitch of PA is that it replaces the vagaries of human intuition and professional experience with hard facts to create evidence-driven human resource management which brings massive efficiencies and benefits for the business.

The effects of the coronavirus pandemic on employment has accelerated the practice and influence of PA. The need for safe distancing at work and expansion of home working has increased the reach of HR and the amount of data available. 

Companies are buying into systems to monitor space usage in real time. Systems such as SPACEWELL can provide maps of office utilisation right down to individual desks. Applications that support home working such as Zoom and Teams can monitor usage and attention, invading spaces that were previously the employee’s private domain.

The consequences of the rise of PA in recent years and the slide of HR into the strategic spotlight are significant for both employees and employers. It is not just that more attention needs to be paid to data protection and privacy since the volume and variety of data used permeates deep into employee activity and behaviour, it is rather that, as HR becomes a centre of strategic innovation the ethical reach and responsibilities expand rapidly.

Using algorithms developed from PA will deeply affect the lives and careers of organisations’ employees. While PA offers new benefits in efficiency and profit for companies, HR will need to keep in mind the importance of employees and how PA could benefit them. 

HR has an ethical responsibility to employees and the danger is that the promotion of HR to the executive board could distract from such responsibility.

Promotions or rotations can be stopped if the analytics suggest so. Machine learning systems which identify behavioural patterns that suggest an employee may move on or take maternity leave (the so-called flight test algorithm) could result in employees being sidelined for promotion or denied training because the investment is not seen as worthwhile. 

Recruitment analytics, drawing on facial recognition algorithms and AI-driven interviews, may exclude the innovative applicant who doesn’t quite fit into the standard structure of projects and organisations. PA has the potential to increase conformity, sidelining the exceptional and dully creativity.

As PA becomes pervasive within organisations, it becomes part of the furniture where employees are no longer aware just how much is being decided by big data and AI. 

Microsoft Viva – released in February 2021 – is an employee experience platform that creates a virtual work experience linking office and home as workplace and incorporates PA and AI. Algorithms can be used to help a manager ‘see if a team is at risk of burnout and then take action to help them develop healthy work habits and feel balanced’. The embedded PA can affect employees’ lives and rights without people realising it.

The effect of PA is pervasive and organisation wide. MS-Viva, for example, promises to amplify culture. But this may mean that abusive and dysfunctional culture becomes even more entrenched. 

Bullying and the demand for conformance and the suppression of dissent can be justified by the evidence-based outputs of PA. The result could be that unethical practice is quickly propagated across the organisation inflicting much greater damage when it’s found out.

In the context of PA and 'black box' technologies such as Hirevue and deep learning models, it is increasingly important for line managers and HR professionals to understand potential pitfalls of technologies and the information they provide. The logical malleability of software increases the need for responsibility by end users. 

Disciplines developed in the research sphere, such as responsible innovation, will be increasingly relevant in guiding how organisations choose and implement new systems, in a way that preserves trust. HR departments in particular may play a key role in equipping practitioners to interpret increasingly complex analytics.

Dr Neil McBride is reader in IT management and Dr Mayen Cunden is lecturer in information systems at the School of Computer Science and Informatics at De Montfort University. Vincent Bryce is a PhD candidate at the Horizon Centre for Doctoral Training at the University of Nottingham