Across the corporate world, whole layers of management are disappearing. GitLab has flattened its management structure by up to three levels, restructuring for what it calls the “agentic era”. Gartner predicts that through 2026, one in five organisations will use AI to flatten their structure, removing more than half of their middle management roles.

It’s been called the Great Flattening. But before you redraw your org chart, there’s a more important question to answer.

Are we seeing a real shift or is this a period of hype spurred on by organisations goals of adopting AI? Right now, it’s difficult to tell but one thing we do know is that the question isn’t whether you can remove a layer of management. It’s whether you should.

When I see announcements like these, the first thing I ask is what investments had to be made upfront to make them possible. Payback justification and business cases need to be robust before you spend one dime.

What is the Great Flattening?

The Great Flattening (sometimes referred to “The Great Unbossing”) is the reduction, or elimination, of the middle layer of management in medium to large organisations. The phenomenon first appeared in the early 2020’s and coincided with the development of AI models which companies have looked to as a way to cut costs.

What’s really driving the Great Flattening?

We all know that it’s AI that’s been getting all the headlines. But I believe it’s cost pressures that are really driving it.

Leaders know AI is very powerful if it’s used properly. But often, the target comes first. A company announces a workforce cut of, say, 20%, and that number then forces disruptive innovation and risk-taking to reach it.

And it won’t happen overnight. Investments need to be made upfront to get there.

You might think this all sounds very familiar, and you’d be right. We were told computers would eliminate middle management in 1958. Reengineering promised the same in the 1990s. We’ve been told twice before that technology would make middle managers obsolete. Both times, the organisations that came out well had a clear business case and brought their people with them, whatever technology was driving the change.

What happens to the work middle managers do?

Here’s the thing: middle managers don’t just sit on an org chart. They run the business day to day. They manage staff, deal with escalations to get product out the door, all while keeping daily management systems running.

I doubt that technology, or the Great Flattening, will get rid of these critical business activities. So when a layer disappears, that work has to go somewhere. That creates two risks:

• The work moves down. It’s very difficult, if not impossible, to remove all of the work of middle management. Some responsibilities percolate down to the next level, and they’re the ones who suffer the impact.

• Strategy gets squeezed. Middle managers run the business so that senior leaders can focus on strategy. Remove the middle layer, and senior leaders risk being pulled into the day to day. It’s a question every leader should ask: will strategic, critical thinking suffer?

What should you ask before you flatten your structure?

If a CEO asked me tomorrow whether they should flatten their structure, these are the five questions I’d ask first:

1. What’s the business case?

2. Is there a clear understanding of what middle management actually does?

3. What will the impact be on the culture of the organisation?

4. Will customers be affected?

5. Has a robust risk assessment been completed?

How many people can one manager realistically lead?

Fewer managers means each one leads more people. So it’s worth knowing where the limit is.

Managing people is all about that one-to-one interaction, be it positive or negative. In my view, the optimal span of control for a typical middle manager is between six and eight direct reports.

Can AI stretch that number? It can, if managers use it to become more productive in their own daily work.

Where does AI help a flatter organisation, and where does it fall short?

The biggest benefit I see is turning data into knowledge. AI can present information in an easily digestible format, so senior leaders can make decisions faster, with the right level of validation behind them.

Where could it fall short? It is very difficult, if not impossible, to remove all of the work of middle management through AI – some responsibilities may need to percolate down to the next level who then suffer the impact.

What does a well-run flattening look like?

When it’s done well, a flatter organisation has:

• Real-time data flowing automatically to senior leaders

• The right level of empowerment at every level to make quick decisions

• Senior leaders spending 80% of their time on strategy and 20% on the base business

• A laser focus on exceptions, where the business is ahead of or behind target

How do you bring managers and teams with you?

Let’s be honest: managers are uneasy about this shift. One survey found 53% of people managers worry they won’t be good at leading teams that work alongside AI.

My advice for managers is to:

• Understand what AI can do to improve your own day-to-day work

• Encourage your team to use AI, within the right boundaries and rules

• “Let AI in”, but with a health warning that it’s not to be taken as gospel

• Help your team strengthen the softer skills that complement AI

The same research found many employees are now reluctant to step into management. So where will future leaders come from? I’m not worried.

AI will not stifle ambition. It’s something new that the workforce will need to adapt to. There will always be ambitious people, and it’s how they adapt to take advantage of AI that will make the difference.

How should you communicate a restructure?

Get this wrong, and you risk spreading fear and losing your best people. The answer is to bring the whole organisation along on the journey:

• Develop a communications plan. Before the first announcement, decide what people need to hear, when, from whom, and through which channels

• Communicate early and effectively, while being mindful that not everything can be shared for legal or confidentiality reasons

• Hold town halls with open Q&A for the whole organisation

• Be open, transparent, and honest, within those legal and confidentiality boundaries

Summary: Quick tips before you restructure

• Be clear on the why. What’s the business case?

• Know which business metrics the restructure will improve.

• Put the customer first. Will it improve performance in their eyes?

• Understand what middle managers actually do before you remove the role.

• Bring your people with you through open, honest communication.

• Don’t restructure for the sake of restructuring.

How can our organisational design team help?

At Pinta, we help organisations review their structures, operating models, and governance. We work alongside your team to plan what’s next, and bring your people with you.

Contact us to start the conversation.