Companies Are Using AI to Cut Managers. They May Be Eliminating the Layer That Makes AI Pay Off

The rise of AI is now leading to a flattening in companies’ operational layers, relying on automated coordination to handle work that was once administered by humans. The average American manager now runs teams of 12 members, up from 8.2 in 2013, as AI accelerates a quiet but seismic shift in how the U.S. workplace is organized. 

 This has been marked as the “megamanager” shift, where companies are using AI to make managers more responsible for more people at the same time under the premise that AI software increases efficiency. 

Regardless, another study by Gallup found that managers account for at least 82% of the variance in employee engagement across business units, which is in turn, responsible for severely low worldwide employee engagement. Only 31% of U.S. employees are actively engaged at work, and a staggeringly low 17% worldwide are engaged; numbers that have remained stagnant over the past 12 years and indicates that the vast majority of employees are failing to develop and contribute meaningfully at their jobs. 

And this issue is mainly due to a gap between employees and the leading bodies of a company; managers are oversaturated with reports and barely have the time and energy to check up on their team. 

If AI removes administrative work, the productivity dividend should theoretically give managers more time for coaching, feedback, communication and relationship-building. But if companies instead respond by giving each manager another 5–10 employees, they may simply convert AI’s efficiency gains into a larger span of control.

“Managers are the ones who connect new processes to the actual relationships and workflows already running on their team. Today as those processes keep changing, skipping that layer will cause a rollout. Build through that layer and you get adoption that holds,” said Darrin Murriner, co-founder and CEO of Cloverleaf. 

AI Can Make Managers Unnecessary – Or Make Them More Important

To reduce costs, companies are racing to use AI and remove layers they once considered essential. 

And it makes sense. A manager who no longer spends half the week compiling reports and chasing status updates genuinely does have capacity to spare, so stretching that role across more people is a reasonable way to book the savings. 

But that calculation overlooks what managers actually do. 

As AI can take over routine work, the human layer that turns projects into realities and turning strategies into employee performance now becomes a strong standpoint for companies. 

If companies remove the layer responsible for translating organizational strategy into employee behavior, where exactly do the productivity gains from AI go?

“That should worry people more than it does, because the relationship with a manager is still the single biggest driver of whether someone stays engaged and stays at the company. Stretch that layer thin instead of investing in it, and you’re not trimming overhead; you’re cutting into the one relationship most likely to keep your best people around and growing,” said Murriner. 

Recent studies have reported on the links between employee engagement at the business-unit level and vital performance indicators, which include higher profitability, productivity and improved quality, as well as lower turnover; less absenteeism and shrinkage. 

This happens when companies choose to invest in proper management strategies, as these really are the root of performance variability and team cohesion. Everyone inside a team has different needs, morals, motivation and clarity, all of which determine employees’ performance and must be supported by a good manager. 

However, good managers are scarce. While knowledge, experience and skills can develop innate talents into strengths, the right innate talents are also critical: without them, no amount of training or experience will necessarily lead to exceptional performance.

A 2023 CMB study suggests that only about one in 10 people possess the high level of talent needed to manage effectively. While many people have some of the necessary traits, few possess the unique combination of talents required to help a team achieve excellence and meaningfully improve a company’s performance. When these individuals are placed in management roles, they are more likely to naturally engage employees and customers, retain top performers, and sustain a culture of high productivity.

“The 10x productivity story people are selling assumes the technology is the bottleneck. It almost never is. The bottleneck is whether the humans around the technology can use it well, and building that capability is what management is for,” said Murriner.

The Hidden Problem with Widening Spans of Control

While companies may initially feel like they are saving money by cutting down management costs, in the long run this might actually result in much more expensive. 

If a manager previously had 7–8 direct reports and now has 15–20, a manager has less time for one-on-one, in-depth coaching, personalized feedback, detecting disengagement, and supporting employees’ career development. 

When there is such a disconnect between leaders and employees, they are much more likely to leave the company, costing between 50% and 200% of that employee’s annual salary, according to Cision Research. 

Recent studies indicate that the average total cost of a single turnover event has risen to $45,236 per worker. 

And it’s not only the costs from employee disengagement, it is also reflected in a blockage between budgets, plans, strategies, and execution. 

“Every dollar you spend on AI still has to pass through a manager before it becomes business value, and cutting managers at the same moment you’re asking them to absorb more people is exactly backwards,” said Murriner

With Cloverleaf, for instance, they condense assessments, personalized coaching, and team insights in one platform so talent leaders can consolidate spend, strengthen teams, and improve performance with personalized guidance. They use AI to gather data, but everything is still handled by humans. 

“What has to stay human is judgment and meaning. Most AI is built to agree with you, and agreement isn’t what a hard moment calls for. Deciding what actually matters when a situation is ambiguous. Delivering hard news in a way someone can actually hear. Sitting with a person through a setback. Those aren’t efficiency problems, and treating them like efficiency problems is how you lose people,” said Murriner.

Similarly, Mariano Jurich, Senior AI Product Leader at Making Sense, a company focused on software development and digital transformation, says the distinction between productivity and business value is often overlooked.

“I think companies get it wrong when they measure output instead of outcome. Output is what you deliver, outcome is what changes because of it. Making a bad process faster is not business value. The fact that a process is like it is, doesn’t necessarily mean that is the way it should be to accomplish business goals and deliver business value at the end of the day.”

The Better Use Case: AI as a Managerial Multiplier

The new dilemma is now more about how we can use AI to improve management, not simply to increase productivity. 

AI can handle routine work; it can do impersonal activities in data analytics, planning schedules, analyze statistics and plan budgets. This frees up time and mental space for managers to handle human-related matters. 

For instance, when managers used to spend around 30 minutes prepping for a meeting, AI synthesizes relevant information and helps the manager prepare much more efficiently. 

Companies are effectively thinking that with AI, they can operate with fewer managers and reduce costs. But a more realistic equation assesses how AI can increase managerial capacity by automating routine tasks, freeing leaders to spend more time on coaching, communication and team development.

This is also showing up in how companies assess their AI investments. For example, platforms like MODO, an enterprise AI platform that measures how AI adoption is changing workflows and ROI investments, demonstrate that many organizations still lack a clear picture of what is actually changing once AI is deployed.

“There’s usually a gap between what leadership assumes AI is doing and what’s happening on the ground: workflows haven’t changed, adoption is low, and no one has visibility into where value is or isn’t being created. Without that baseline, it’s very hard to know whether the investment is paying off,” Johnny Chang, co-founder and CEO at Modo told Unite.AI. 

Essentially, teams do not need more systems or elegant software for their own sake; they need support and motivation from leaders who can help them adapt, collaborate and ultimately contribute to the company’s growth.

Platforms like Cloverleaf are designed to make that kind of management more scalable, using AI to give managers personalized insights and coaching while keeping the human relationship at the center of the process.

“We studied nearly 30,000 real coaching conversations, and in half of them the employee named a specific colleague, not a skill or a competency. They wanted help with a person. They also asked to understand a colleague 2.5 times more often than to resolve a conflict, which tells you people will do the relational work early if you give them the support to do it. That’s the behavior change we’re after, and it’s measurable,” said Murriner.

The takeaway is not that AI has no place in management, but that its greatest value may come from strengthening the people already responsible for leading teams. Companies that treat managers as a cost to eliminate risk cutting away the very layer that turns AI-driven efficiency into sustained performance.

“The pattern underneath all of it is the same. When you strengthen the manager instead of replacing them, people engage with development, they do the relational work sooner, and the team performs better. None of that happens when you treat management as a cost to cut,” said Murriner.