Algorithmic Management – Have We Dehumanized the Workforce?… When Efficiency Gets Measured by the Minute, Where Does Empathy Fit In?

A Day at Work, According to the Algorithm

8:57 AM – Before the First “Good Morning”

At 8:57 a.m., an employee logs in and finds that much of the day has already been mapped out. Tasks are waiting, targets are visible, and priorities have been set. A manager may not have said a word, yet technology is already directing the rhythm of work. This is where algorithmic management enters the picture: software and data-driven systems are being used more to allocate work, monitor employees, evaluate performance, and support managerial decisions. The OECD notes that these systems are changing how work is organized and how employees are managed. 

The appeal is obvious. Businesses want faster decisions, better coordination, and greater efficiency. But when more workplace decisions move onto screens, an uncomfortable question emerges: when technology gets better at measuring employees, are organizations becoming worse at understanding them?

10:32 AM – The Appeal of the Perfect Number

By 10:32, the employee has already generated a trail of numbers: tasks completed, response times, customer feedback, hours worked, and progress against targets. For managers handling large or distributed teams, algorithmic management can provide a practical way to process this information, identify patterns, and support decisions around scheduling and performance. The OECD notes that managers value them partly because they provide information that can support decision-making. 

There are clear practical benefits. AIHR highlights how algorithmic systems can automate repetitive management tasks and improve efficiency. Its discussion of L’Oréal, for instance, describes the company’s reported use of an AI recruitment solution to speed up hiring while allowing more applicants to be interviewed. 

But a number only tells part of the story. A response time can be measured, but it cannot always explain why an employee took longer or what happened during that interaction. A number can tell a company what happened; it may not tell the company why.

12:46 PM – The Twenty Minutes the Dashboard Missed

At 12:46 p.m., an employee stops to help a colleague with a difficult task. Twenty minutes disappear from the schedule. On a productivity dashboard, those minutes may look like lost output. In reality, a colleague has learned something, a mistake may have been prevented, and knowledge has moved through the team.

This is where the debate around algorithmic management becomes a question of values. Research has raised concerns about how automated systems can change workplace control, employee autonomy, and working conditions, while also stressing that their impact depends on how they are designed and governed. 

Some of the most important contributions at work are difficult to measure. Mentoring, patience, informal leadership, and helping a struggling colleague rarely appear neatly on a performance dashboard. Yet they can determine whether a team actually works well.

3:15 PM – When the Dashboard Speaks First

At 3:15 p.m., a manager receives an alert: an employee’s performance has fallen below the expected threshold. The manager can ask, “What happened?” or begin with, “Why are you below target?” The difference may seem small, but it reveals how easily management can shift from conversation to measurement.

These systems can help managers spot changes they might otherwise miss, particularly across large workforces. The OECD identifies decision support as one of their potential advantages. 

Still, the dashboard does not know everything. It may not know that the employee was dealing with an unusually difficult customer, covering for a colleague, or solving a problem that never entered the system. Technology can provide the starting point for a conversation, but it cannot always replace the conversation.

5:41 PM – Learning to Please the Algorithm

By late afternoon, the employee has learned what the system rewards. If speed matters, work becomes faster. If volume matters, quantity takes priority. If customer ratings influence performance, difficult interactions may become something to avoid. Employees do not simply work within systems; they learn how those systems work and adjust accordingly.

Research on algorithmic management examines this relationship between automated control and worker behaviour, including how people adapt to systems that influence their working conditions. 

That creates a subtle problem. An organization may believe it is measuring performance, while employees are learning how to perform for the measurement. The result can be better numbers without necessarily producing better work.

6:30 PM – Everything the Algorithm Doesn’t Know

By the end of the day, the system may know an extraordinary amount about an employee’s activity. It can record tasks, timings, output, and ratings. Yet it may not know whether that employee is exhausted, losing motivation, feeling undervalued, or quietly considering leaving.

The concern becomes greater when monitoring turns continuous. These issues sit at the heart of the debate around algorithmic management, particularly as Esade examines the pressure workers can experience when AI is used to monitor and evaluate them, while Lewis Silkin discusses workplace surveillance and the creation of a “culture of proof,” where employees feel more compelled to demonstrate that they are working. 

A system can capture an employee’s activity without ever capturing the circumstances behind it. An employee can be measured constantly and still feel unseen.

The Next Morning – The Algorithm Is Not the Villain

Blaming the technology would make the debate too easy. Businesses have legitimate reasons to use these systems. A large organization cannot rely entirely on individual observation, and managers cannot manually process every piece of information available to them.

The OECD recognizes both the opportunities and risks of algorithmic management, emphasizing that much depends on how these systems are implemented and governed. 

There is also a case for using technology to strengthen, rather than weaken, human connection. Cognizant explores how AI can support empathy by helping managers understand employee needs, reducing administrative work, and creating more room for meaningful interaction. So perhaps the question is not whether algorithms belong in management, but where their role should end.

The Workplace We Choose

The workplace is unlikely to become less dependent on data. The real choice is whether organizations use that data to understand people better or simply to watch them more closely.

The future of algorithmic management will depend on how thoughtfully these systems are used. Technology can process information, identify patterns, coordinate schedules, and flag changes. People still need to handle what requires context: listening, mentoring, questioning, judgment, and empathy. That means being transparent about how systems are used, allowing employees to understand decisions that affect them, and maintaining human oversight when those decisions carry real consequences. 

Perhaps the better future for algorithmic management is not one without algorithms. It is one where the algorithm knows its place. A productivity alert should begin a conversation, not end one. A performance score should prompt a question, not become a verdict.

The employee who logs in at 8:57 a.m. is more than the figures appearing on a screen. The danger is not that machines are learning to manage people; it is that, as they become better at measuring people, we may become worse at understanding them.