Once upon a time, organizations were made up of people. Today they consist of data. As companies have learned to mine their data to better identify new opportunities, improve predictions, and make better decisions, interest has shifted from the humans who do the work to data on what they do during work hours (e.g., how may emails they sent, how many people they talked to, how many breaks they took). In particular, employee data is being used more and more in human resources management (HRM) — and, more recently, people analytics (PA) — and workers are increasingly being defined in terms of their data.
The implications of this shift are significant. An approach that defines people and their value to the company (actual and predicted) in terms of data runs the risk of depersonalizing the people that make up companies, reducing them in the eyes of their employer to the level of interchangeable objects. Moreover, it has the potential to create a work culture that denies employees’ privacy and in which people feel less safe.
This trend of depersonalizing employees isn’t necessarily new. For some time now, HRM has focused less on approaching the employee as a “whole” human being and more on promoting a one-size-fits all approach to manage employees. To reduce costs and promote efficiency of compliance and standardization, HRM has approached employees mainly in terms of the quotas they are supposed to hit, the sales they make, the deals they close, and so forth. There’s an ugly logic to this: treating employees as interchangeable commodities makes it easier to impose the ever-increasing burden of bureaucracy that defines contemporary organizations.
Recently, however, this approach has evolved: human resource management has paved the way for people analytics, which uses statistical methods and intelligent technologies (e.g., sensors, digital devices) to create and analyze digital records of employee behavior and employ an evidence-based approach to increase the organization’s efficiency and productivity. Today, up to 70% of executives consider the implementation of PA capabilities as a top priority and predictions are that the value of the global big data analytics market will be around $68 billion by 2025.
PA goes beyond the traditional procedures for measuring and quantifying employee performance — intelligent technologies working with large, unstructured, real-time data and aggregated data sets allow organizations to make predictions rather than simply measure outputs. But, the real departure from traditional HRM practices is that PA often means that employees are surveilled and analyzed at increasingly intimate levels all the time. Data from devices such as cameras, Bluetooth beacons, mobile phones, IoT wearable devices, and environmental sensors are analyzed with the aim of making predictions that allow supervisors to address, evaluate, and — if needed — punish employee behavior. This situation has resulted, for example, in companies firing employees when they find out that those employees are applying for other jobs. The European Commission found that global demand for employee-spying software more than doubled between April 2019 and April 2020 and that during the lockdown period in 2020 surveillance software tracking time that employees actually spend on their tasks quadrupled.
This monitoring comes at a price, however. The creep of surveillance deeper into more and more parts of the workday means that employee privacy is all but being erased, and their work experience is being negatively affected. Fear that someone is always looking over your (figurative) shoulder undermines trust; monitoring can hurt employee morale and actually make people act less ethically. In other words, surveillance can produce the opposite of the intended effect and create a work environment that runs counter to contemporary recommendations from coaches, consultants, and trainers that organizations today need to develop a culture that empowers people, rather than returning to central control and rigid processes.
What to Consider When Implementing People Analytics
Introducing and reinforcing a mindset where people are reduced to their data can create a work culture that may do more harm to performance and employee experience than anticipated. It is therefore important that organizations adopting PA do not turn HR departments into IT departments focused on monitoring and optimizing workers’ efficiency, and rather ensure they safeguard employee interests in empowering ways. To do so, below, we briefly discuss three strategies that the organization of today takes care of to ensure PA is used and seen as empowering human employees.
1. Make clear it’s not a step towards automation.
PA comes from organizations’ growing use of artificial intelligence (AI) to promote efficiency. An inherent risk here exists, however: this operational way of working may promote the idea that people are secondary to the systems that monitor them, leading employees to feel that they’re simply producing data for their AI bosses, and that they’re increasingly replacable. This can instill the notion that PA is a step towards automation, in which people are training the machines that will replace them. Rather, when PA is implemented, it should be framed and instituted as a strategy that augments the abilities and performances of employees, thereby indicating that humans are first priority and machines only secondary.
2. Realize that people analytics is about more than efficiency and show it.
Communicate clearly that PA will not only be used to predict individual employees’ performance — an approach that errodes trust and infringes on employee privacy. Organizations should avoid framing performance as the end in itself, which communicates that employees are simply a means to achieve that end. Instead, companies should focus on how monitoring and analysis can be used in a holistic — rather than narrow — way to help employees grow and develop as people, and emphasize that focus in communications to workers. For example, data can be collected to identify where employees experience stressors and used by the organization to offer help to deal with obstacles of personal growth. Indeed, tracking the sentiments of employees — in anonymous ways — and identifying significant trends can help to promote more empathy from those in charge while at the same time enhancing a common understanding of the stressors that are experienced at work.
3. Avoid labeling people as data.
Using PA successfully to motivate employees also depends on the kind of narrative that is being used. For example, when cc-ing employees in emails sent by HR to supervisors, avoid abstract language where the employee is regarded as a number or described in depersonalized ways. Describing people as data, company assets, or investments that need to reveal some ROI, communicates that an employee is seen as an object and in a way thus not deserving of humane attention to their own development and growth. Do not forget that people bring their entire selves to the workplace, and as such they will be responsive to the extent that the organization appreciates who they truly are.
Taken together, in an era where organizations are data-driven and intelligent technologies facilitate the tracking and assessment of employees in a variety of ways, it is more important than ever that humans feel included, valued, and treated in empowering ways.
Organizations that adopt PA strategies where employees are primarily identified as datasets to enhance transparency about whether everyone is doing their job run the risk to create an empathy gap where employees will feel — despite the abundance of personal data collected — poorly understood. Under those circumstances, PA will likely be seen as being more in service of creating a face-less workforce that acts as efficient as machines are expected to be rather than creating facilitating conditions under which employees feel that the organization invests in their growth and self-development. It is therefore necessary to realize that collecting and analyzing employees’ data can be useful and valuable to the organization, but not if it primarily makes employees feel like quantifiable objects in a machine-driven context.