Your next hire has no salary and no ceiling

Open any company’s payroll and you’re looking at a list of ceilings. Each name is a person, and each person can do roughly one person’s worth of work in a day, no matter how good they are, no matter how much you need more of them by Friday. You pay for that ceiling every month, in full, whether the week was slammed or dead quiet. That is what a hire has always been: a fixed cost bolted to a fixed throughput, decided the day you sign the offer.

I think we’re about to stop hiring ceilings. The unit of capacity is changing underneath us, and most org charts haven’t noticed yet.

The new unit of capacity is a role you switch on, and the only cost that scales is the work it actually does.

The thing a hire actually is

Forget the résumé for a second and look at hiring as a transaction. You are buying capacity, the ability to get a category of work done, and you are buying it in a very specific, very old shape.

You buy it as a whole person. The person comes with a fixed maximum output: one human’s hours, one human’s focus, one human’s hands on one keyboard. You pay a flat rate for that maximum whether or not the work is there to fill it. And you can’t buy a fraction of it on a Tuesday and give it back on a Wednesday. The capacity is lumpy, it’s permanent, and its price has almost nothing to do with how much work shows up.

A hire is a fixed cost stapled to a fixed ceiling. You pay for the ceiling, not the work.

For the whole history of business this was simply the only shape capacity came in, so nobody questioned it. If you needed more qualified leads, more drafted contracts, more reconciled invoices, the only lever was another person, with all the lumpiness that drags along: the search, the months of ramp, the floor under their cost that exists whether the quarter is busy or slow.

The new unit of capacity is a role you switch on, and the only cost that scales is the work it actually does.

The naive fix: rent the person by the hour

The obvious patch is to make the person more liquid. Contractors, agencies, fractional hires, gig platforms, the whole industry of capacity on demand exists to file the lumps off the old unit.

And it helps, a little. You can scale a contractor up for a busy month and cut them after. But look closely at what you actually bought and the old shape is still in there, intact. You’re still renting a whole human’s ceiling, just for a shorter lease. The contractor still does one person’s work in a day. They still need to be found, briefed, and ramped before they’re useful, so “on demand” really means “on demand in two to three weeks.” And the moment you need ten of them next Friday, you’re back in a hiring cycle, a faster one, but the same shape.

The naive fix treats lumpiness as a scheduling problem. It isn’t. Lumpiness is baked into the unit. As long as the unit of capacity is a whole person, you inherit the person’s ceiling and the person’s ramp no matter how cleverly you lease them. You can make the person cheaper to start and cheaper to stop. You cannot make one person do four people’s work on the Friday you suddenly need it.

So the question isn’t how to lease the old unit more flexibly. It’s whether the unit itself can change.

Our way: hire the role, not the head

Here’s the shift, and it’s smaller than it sounds. Stop buying capacity as a person. Start buying it as a role, a precise description of a job to be done, that you can switch on, run as wide as the work demands, and pay for only by the work it does.

A role is not a smaller person. It’s a different kind of thing entirely. You define it once: qualify a lead against these criteria, draft a first-pass contract from this template, reconcile these statements and flag the mismatches. Then you switch it on. There’s no search and no three-month ramp, because the role is the spec, and the spec is ready the moment you’ve written it down clearly. That part, writing down what “good” looks like, turns out to be the only real work left, and it’s work you should have been doing for your human hires anyway.

The two properties in the title fall straight out of this.

No salary: cost meters on work, not on time

A salaried person costs the same in a dead week and a brutal one. The cost is attached to the existence of the capacity, not its use. That’s why a quiet month feels like waste and a slammed month feels like you’re short, the price never tracks the work.

A role inverts that. It costs nothing while idle, because there’s nothing to keep warm, no chair, no retainer, no floor under the bill. When 400 leads land Friday morning, the role runs 400 times and you pay for 400 qualifications. When Monday is quiet, it runs four times and you pay for four. The cost is a function of the work itself, the way your cloud bill tracks the traffic you actually served rather than the servers you might have needed. Capacity stops being a standing expense and becomes a variable one, metered to the thing you were trying to buy all along.

No ceiling: scale by running the role wide

The deeper property is the ceiling, and it’s the one that breaks the old intuition hardest.

A person’s output is capped at one person. To get more, you add more people, slowly, lumpily, one hire at a time. A role has no such cap, because the way you scale it is not “add another role,” it’s “run the same role wider.” The lead-qualifier you defined once can run as one instance or as fifty at the same moment, against the same criteria, all on the Friday the backlog spiked. You didn’t hire fifty qualifiers. You switched on the one you already defined, fifty times over, and switched it back to one when the spike passed.

That’s the line that doesn’t exist for a human hire. One person is one person’s ceiling. One role is whatever throughput the work demands, because the throughput is rented, not owned, and you give it back the instant you’re done.

You don’t hire fifty. You define one, run it fifty wide, and scale it back to one when the spike clears.

Where the old org chart quietly breaks

This is the part that sounds abstract until it lands on a real team, so make it concrete.

Suppose a small firm’s whole growth depends on qualifying inbound leads, and a campaign lands 400 of them in a morning. Under the old unit, that firm has exactly two moves, both bad. Burn out the two people who do qualification, pushing the work into next week and losing the leads that go cold. Or staff for the spike year-round, pay for ten people’s ceiling every month to survive the few mornings it’s actually needed. Lumpy capacity forces you to choose between too slow on the peak and too expensive on the average, forever. Every founder has felt this vise. It’s not a management failure; it’s the unit.

Switch the unit and the vise disappears. The qualifier role runs 400 wide that morning and back to a trickle by lunch. Nobody burned out, nobody got paid to wait. The firm sized its capacity to the work that day, not to the worst day it might ever see. That flexibility was never available before, because you can’t fractionally summon and dismiss a person between breakfast and noon. You can do exactly that with a role.

Notice what this does to the org chart. The old chart is a map of fixed ceilings, boxes, each one a person, each one a permanent cost and a hard cap. The new chart is a map of roles you can switch on, most of them dark most of the time, lighting up wide when the work spikes and going quiet when it passes. The headcount stops being the measure of what the company can do. What it can do is measured by which roles it has defined well, and a defined role has no ceiling that a busy Friday can hit.

The turn: defining the role is the human’s job

So if the roles do the work, what’s left for the people?

The most important thing in the whole system, it turns out. A role is only as good as the spec behind it, and writing that spec, deciding what “a qualified lead” actually means for this business, what a good first-pass contract looks like, where the line sits between “flag it” and “fix it”, is judgment work that no role can do for itself. The machine can run a definition a thousand times in parallel without tiring. It cannot decide what the definition should be. That decision is taste, and context, and knowing your customers, and it stays with the humans because it was always the part that mattered.

This is the quiet inversion. For most of business history, people spent their days being the capacity, doing the qualifying, the drafting, the reconciling, one item at a time, until the day ran out. When capacity becomes a role you switch on, people stop being the throughput and start being the thing that decides what throughput is for. You move up a level: from doing the hundred tasks to defining the one role that does them, from filling the ceiling to deciding which work deserves a role at all. The least replaceable thing a person brings to a company was never their hands on the keyboard. It was knowing what good looks like, and that part doesn’t fan out, and shouldn’t.

---

That’s what we’re building at Apollo Space: a company where capacity is a role you switch on, priced to the work and unbounded by headcount, so the people are free to do the one thing a role can’t, decide what’s worth doing in the first place. If you’ve ever stared at a spiked backlog and a payroll full of ceilings and felt the vise close, there’s a different shape on the way. Your next hire doesn’t need a desk. It needs you to tell it, clearly, what good looks like.