How the rise of AI is putting flat management structures to the test

Will hi-tech firms adopt more hierarchical, authoritarian structures to ensure that their staff will keep working in the best interests of the business rather than protecting their jobs against automation?
Businesswoman leading a meeting with her peers

Hi-tech firms have been among the most enthusiastic adopters of ultra-flat management models such as holacracy, whereby fluid teams decide what work must be done, who will do it and how their efforts will be rewarded.

The underlying idea of this approach is that a conventional hierarchy would only stifle the output of highly educated tech workers, who have an intrinsic interest in their occupation and are perfectly able to self-manage. These people aren’t doing menial work; they’ve been hired for their creative problem-solving skills. It would be impossible for a management team lacking their technical knowledge to understand every aspect of the complex work they do. Indeed, many such professionals would probably resent being subjected to top-down management and behave counterproductively as a result.

What happens to that intrinsic interest, though, when automation requires them to start designing away a significant proportion of their work? ChatGPT has hit the headlines because its responses can seem sentient, for instance. The technology has learnt to perform some tasks that, until now, only people could do. 

The shift towards automation could make self-management more compelling

But this quantum leap in artificial intelligence is not necessarily the biggest star in the automation show. For instance, AI-assisted robots have made a breakthrough in what was previously a ‘bottleneck skill’ – that is, something machines couldn’t do. The first few generations of robots, such as those on production lines, performed only programmed tasks in completely predictable environments. But the latest ones can handle objects of varying dimensions in unpredictable orientations with ease. 

A study published by the OECD in December 2022 estimates that, on average, “occupations at highest risk of automation account for about 28% of employment” in its member nations. Its research report states that “only” 18% to 27% of the abilities required by high-skilled occupations are “highly automatable”. But that still suggests that a significant percentage of what some highly skilled people currently do will one day be done by a machine. 

Could that create any tensions in organisations that place a high value on self-management? And might they need to change their approach as the march of automation continues?

Why tech firms won’t crack down on self-management

The most innovative tech companies are unlikely to grab back the reins, according to Michael Y Lee, assistant professor of organisational behaviour at the Insead business school. These firms typically seek out the most talented people and make the most of that talent, so a flat, decentralised approach will normally suit their requirements, he argues. 

“Research shows that the more hierarchical a firm is, the less innovative it will be. Decentralised firms, by contrast, tend to miss fewer of the good left-field ideas,” Lee says.

But a decentralised approach does have its risks. Such organisations tend to be harder to lead and, while they offer a conducive environment for ideation, that’s likely to produce more bad ideas than average as well as more good ones. 

“As tech businesses move towards automated processes to streamline inefficiencies, there’s still a human element that needs steadfast management,” says Ralph Dangelmaier, global CEO of BlueSnap, a global payments company. “The responsibility is on managers to then educate their teams as to why this enhances their job, rather than replaces it.”

The self-management model has the potential to benefit any business, not just those employing the intellectual crème de la crème. For instance, a landmark 1949 study published by the Tavistock Institute of Human Relations revealed how a group of miners in Yorkshire had boosted the productivity of their colliery by changing the way they worked. And, some 40 years later, US firm C&S Wholesale Grocers also showed how letting a group work out how best to meet certain targets could greatly improve outcomes.

“If you shift decision-making down, you get more responsiveness at the local level and the edges,” Lee says. “The flipside is that a self-organised firm may take longer to mobilise support for a major shift.” 

Can decentralised firms handle big layoffs?

And a “major shift” is exactly what big tech firms are having to handle. But the huge number of redundancies across the tech sector in recent months – Google laid off 12,000 people in one round, while Meta cut 11,000 jobs – may not be as damaging to morale within these companies as outsiders might imagine.

“Decentralised organisations tend to have stronger cultures, which is partly because of self-selection during recruitment,” Lee says. “Those people probably also buy into the argument that only the strongest performers should be on board, so they may accept layoffs.” 

Self-management is a bit like the road to enlightenment. There is no destination

He points out that there are, in any case, two types of self-management. One is more libertarian, placing the greatest value on autonomy and freedom. The other is more cohesive and community-based, valuing citizenship and reciprocity. And, while firms in the latter category tend to be more reluctant to make redundancies, it’s not true that a big round of layoffs will necessarily turn those remaining in the business against automation.

Dangelmaier argues automation has benefits that staff will embrace, especially where it removes arduous, repetitive tasks. “In the accounts receivable process, one in four employees has reported an increase in morale under a management style based on automation. Increasing automation does not necessarily conflict with team goals – instead, it uplifts teams.” 

Why automation means more self-management

“The shift towards automation could make self-management more compelling,” Lee argues. “It should remove low-knowledge work and leave high-level tasks to humans. That will raise the bar of what it takes to be employed.”

This in turn would leave management teams having to deal with organisations full of experts. Such companies would move even further away from a hierarchy, according to Lee. He suggests that this will require business leaders to focus on getting people with specialised expertise to cooperate effectively. 

“It’s never just one person who creates an innovative product. An effective, collaborative team is where the magic happens,” Lee says “How you design the team, choose the right people and create direction is what matters. The design element tends to be neglected in management training.”

Warning that the natural place for all of us to be is in a hierarchy, he adds: “Self-management is a bit like the road to enlightenment. There is no destination. It’s a continuous process of gaining greater capacity, flexibility and skill.”