Hiring for Headcount Was Always a Lie. Agents Make It Obvious.

Growth does weird things to companies.

Early on, everyone is fighting for relevance. You're trying to find customers, ship something useful, and not get crushed.

Then, if you're lucky, the company starts working. And the game changes.

At some point, growth stops being only about the customer. It starts becoming internal. Who owns what? Who gets the headcount? Who has the larger team? Who has the bigger scope? Who gets invited to the meeting before the meeting?

The work is still there and the customer still matters. But a new scoreboard appears.

Headcount.

I remember one year at Apple where the target was to roughly triple the size of our team. The logic was simple enough: we had too much work, so we needed more people.

That is often the case.

But something else happens when the number gets big enough. Team size becomes a status symbol. The larger your team, the more important you must be. The more people reporting into you, the more serious your work must be. The org chart starts to look like a scoreboard.

That is the part agents are about to make uncomfortable.

People are not a currency.

That sounds obvious until you’ve worked inside a large company for more than ten minutes.

A team’s importance is not measured by its size. A manager’s value is not measured by how many people report to them. A company does not become more capable just because the org chart gets taller and wider.

At times it does, but only when the people map to useful work.

That is the part that gets blurred.

Companies need teams because the sum can be greater than the parts. A great team is not just a pile of talented individuals. It has shape. It has chemistry. It has roles. It has trust. It has an understanding of what each person is there to do.

A team exists to serve the work. Not the other way around.

I spent a year at Wharton on human capital: how people, teams, and structures actually move a business. The org chart grows because the promotion ladder grows. Output is the residual.

Good individual contributor becomes manager. Good manager becomes senior manager. Good senior manager becomes director. Director becomes VP. You get the picture.

There is nothing inherently wrong with this. Some people become better as they move up. Some people are meant to coach, structure, hire, clarify, and raise the quality of a group.

But the ladder has a flaw.

We often promote people because they were good at the job they are leaving, not because they are good at the job they are entering.

That is the Peter Principle in plain English: keep promoting people for being good at the last thing, and eventually they land in a role where they are no longer good.

Then the organization solves the wrong problem. It adds more people.

More coordinators. More managers. More meetings. More reviews.

Now some of this is necessary. A world-class researcher? Hire them. Someone to keep production alive at 3 a.m.? Please hire them. A specialist who can solve a problem nobody else understands? Worth it.

A lot of headcount isn't about specialization, and it's simply organizational overhead.

This is where agents get interesting. And awkward.

I run a company as a solo founder, now five agents, and roughly 30 PRs a week. Not a humblebrag, I’m actively testing the new world order. When the work is clear, agents move. When the work is muddy, they drift.

Give someone coding agents, research agents, scheduling agents, QA agents, and a clear goal, and the questions change.

“How many people do we need?” “What work actually needs to happen?”

That is a massive shift. Because once you ask the question that way, a lot of organizational language starts to look suspicious.

Drive alignment. Own the process. Support cross-functional visibility.

Some of that work matters. Some even matters a lot. But much of it exists because the work was never defined clearly enough in the first place.

People can survive inside ambiguity. In fact, many people build careers inside ambiguity. They know how to read the room, navigate politics, patch broken processes, translate between teams, and keep the machine moving.

That is a real skill. It’s also expensive overhead.

Agents do not do well with vibes. Vibes don't compile. They do not know that Jess in finance hates being surprised on Fridays. They do not know that the VP says “looks good” when he actually means “I have concerns but don’t want to say them in this meeting.” They do not understand that the roadmap is technically approved but socially dead.

Humans are still better at all of that.

But agents are very good at making the work legible.

Give an agent a clear task, boundary, input, output, and definition of done, and it can move. Give it a mushy job description and a political environment, and it falls apart.

That is the tell.

If a role cannot be explained clearly enough for a capable agent to help with it, there are only a few possibilities.

The work is deeply human. The work is poorly defined. Or the work is mostly for show.

The hard question for every leader now is simple:

If you rebuilt the company from scratch today, would you design it this way?

Same people. Same customer. Same market. Same goals.

But now everyone has access to agents. Would the org chart look the same?

If you’re honest, probably not.

That does not mean “fire everyone.” That is the lazy version of the take, and also the cruel one. Most companies are full of good people trapped inside bad work design.

Define the outcome first. Then the workflow. Then the tasks. Then decide what should be done by a human, what should be done by an agent, and what should disappear entirely.

Some work needs judgment. Some needs taste. Some just needs to be done.

We have historically hired people for all of these. That era is ending.

The next org chart will probably look less like a pyramid and more like a small core of high-judgment operators, a set of specialists, and a lot of agents doing the work that used to disappear into process.

Using AI is easy. Redesigning work around it is so much harder.

The political part of work has always been there. Agents make it easier to see.

When work is vague, politics fills the gap. When roles are unclear, status fills the gap. When outcomes are hard to measure, headcount fills the gap.

And companies drift.

A little more scope here. A little more coordination there. A new weekly review. A new planning layer. A new person to make sure everyone knows what everyone else is doing.

Then one day someone asks what all of this is for, and the room gets quiet. Agents are going to force more of those rooms.

That will be painful and honestly it should be. It’s also healthy.

The best people do not want to spend their lives maintaining corporate fog. They want to do work that matters. They want clear ownership. They want fewer fake deadlines, fewer status games, fewer meetings where everyone performs certainty and nobody says the thing.

So the question is not whether your company needs fewer people or more agents.

The existential question becomes:

And if everyone in your org became a free agent tomorrow, which roles would you design again?

What if SEO ran without you? Over 2,500 AutoSEO customers already show up in Google, ChatGPT, and Perplexity. Expert articles, backlinks, mentions — written and earned after setup without you touching it. No learning curve. No weekend blog posts. No agency. Get a free plan built around your business in 3 minutes.