I sat in a meeting last month where a leader said, "we don't need managers anymore, we need player-coaches." I wrote it down because it sounded sharp. Then as I thought about it, I wondered who was going to coach the next generation of player-coaches.
That's a significant problem with what's happening in org charts right now; something I'm not sure is being weighed carefully when axing these managers.
The flattening is real, and it's not slowing down
Gartner's forecast: by the end of this year, one in five organizations will use AI to flatten their structure, cutting more than half their middle management layer. Amazon cut 14,000 corporate roles last fall in the name of "removing organizational layers." Block eliminated roughly 4,000 of its 10,000 employees in February, with Jack Dorsey framing it as permanent, moving to a three-role structure of individual contributors, directly responsible individuals, and player-coaches. Nobody's hiding this anymore. It's the strategy.
And it's not just the giants. Bessemer Venture Partners just reported that 49% of their own AI-native portfolio companies say they're already delivering more output without adding headcount. Their talent team's read: it's not that jobs are disappearing, it's that the roles themselves are blurring together. Finance leaders are expected to model AI-native unit economics. Engineering leaders are spending more time with customers. Marketing leaders are getting more technical. Every function is quietly absorbing what used to be three people's jobs into one seat.
And look, some of it is overdue. A lot of management layers existed to shuffle status updates that AI now does in ten seconds. Nobody's mourning that.
Here's the part getting skipped
Middle management was never just a reporting layer. It's where a company grows its next senior leaders. It's where a 30-year-old gets her first shot at owning a P&L, running a team through a bad quarter, making a call that's actually hers to make. Cut that layer out entirely and you're not just saving payroll. You're eliminating the training ground.
One Forbes piece put it bluntly this spring: the companies eliminating their middle managers in 2026 are quietly dismantling the exact system that would have produced their senior leaders in 2028. That's not a talent problem you can hire your way out of on short notice. Judgment takes years to build. You can't rush-order it.
So here's the bind a lot of leaders are in right now, whether you're a founder scaling fast, a company mid-integration after an acquisition, or a PE firm two years into a hold: you need senior-level judgment sitting in the room today, but you just flattened the structure that used to grow it internally, and you're not ready to make a permanent multi-year bet on rebuilding that layer from scratch.
The market already figured out the answer
While headcount at the manager level keeps shrinking, demand for fractional and interim executives is doing the opposite. 77% of business leaders say AI is increasing their need for specialized outside talent, not reducing it. 72% of CEOs say they plan to use more fractional executives in the next year, not fewer.
And this isn't a big-company phenomenon parked at the enterprise level. Heidrick & Struggles' 2026 Talent Lens survey found 42% of interim leaders work mostly with companies under $50M in revenue, another 41% with companies between $51M and $1B. The single biggest reason companies bring someone in isn't a skills gap. It's a transition: a growth stage, a post-acquisition integration, a leadership change that needs an experienced hand in the chair before anyone commits to a permanent decision.
That's not a consolation prize for companies that can't afford a full-time hire. That's the correct structure for a moment where org charts are actively being rebuilt and nobody wants to guess wrong.
What this means if you're the one flattening
If you're cutting layers right now, you're not wrong to want speed and lower fixed cost. Just be honest about what you're actually trading away, and don't try to replace judgment with a dashboard.
Oracle just bet $2.8 billion on cutting Customer Success and letting agents hold the retention relationship.
Telstra made the opposite bet the same week: same agentic tools, but in front of 1,000-plus reps, not instead of them, automating the admin so people can handle what's actually complex. Only 9% of customer conversations get fully handled by AI without a human right now.
Bet on which model ages better. Bring in someone who's already run the play, on a timeline that matches how long you actually need them, not a permanent headcount commitment you'll regret rebuilding around in eighteen months.
Not sure what you need — a hire, a fractional pro, or just a second opinion? That's exactly the call we help founders and business leaders make.

