The $1M-Per-Head Firm

The number Marcus had never asked for

Marcus asked Claudia (CFO/COO) to run a number he had never given serious attention to before. Revenue per corporate FTE.

Fund-level returns. Deal-level IRR. Management fee waterfall. LP distribution schedule. He had been staring at those numbers for twelve years. They told him about the portfolio and profitability, while leaving the firm's top-line efficiency invisible.

Claudia pulled it in an afternoon. A series of fund asset management fees across three active funds, plus a pool of acquisition, financing, and disposition fees as well as service fees including construction management and development fees. The denominator included twenty-two corporate FTE in the Austin office, including Marcus. The operating-partner personnel at properties (roughly one hundred and eighty across the portfolio) are paid out of property-level operations and do not sit on Chen Capital's payroll.

The result: $682,000 per corporate FTE.

Marcus looked at the number and felt fine about it. Mid-market real estate private equity firm, twelve years in, growing AUM, solid returns. $682,000 per head seemed reasonable. It would not seem reasonable for long.

Claudia had pulled one more thing, unprompted. She had taken the Coordination Tax estimate from Marcus's napkin math, forty percent of productive capacity consumed by assembly, reformatting, and reconciliation, and mapped it against the team's weekly hours.

"The $682,000 is real," Claudia said. "But as we've realized, forty percent of the team's time is going to assembly work. They are operating at maybe half their capacity. The tax is capping how much AUM this team can manage rather than reducing our revenue. We are leaving growth on the table."

Marcus asked the question that had been forming since the Driskill. "If a buyer looked at us right now, really looked, what would they pay?"

Claudia had run a version of it already. "Less than you would want. Start with what recurs. The management and asset-management fees are real, but after what it costs to run this place, all twenty-two of us, most of our hours going to assembly, what is left as durable earnings is thin. Five, maybe five and a half million in a normal year. The acquisition and disposition fees swing with the deal calendar, so a buyer discounts those. And the promote." She paused. "The promote is the number we toast at the holiday party. Lumpy, years out, contingent on exits. A buyer pays almost nothing for it today."

"So they capitalize the thin number."

"They capitalize the thin number, then take the founder discount off the top, because the day you and David walk out the earnings walk too. A bit over three times a figure I would not put on a slide. Call it the high teens of millions. Seventeen, give or take. Most of what you think the firm is worth does not survive the diligence."

Marcus sat with it. "Twelve years, three funds, competitive returns, and we are worth three times a number I am almost embarrassed by."

"We are not a company yet. We are a well-paid group of people, and the market can tell the difference." She tapped the printout. "The revenue-per-head number and the valuation number are the same story told twice. Thin fees per person, capped by the tax, dependent on the people, that is a low multiple. Lift the fees each person can carry, make them recur, make them run without you, and you have not just built a more efficient firm. You have built a more valuable one. A sellable one."

That last phrase was the one Marcus carried into the afternoon.

---

Same team, twice the portfolio

Marcus brought David Kwon (senior asset manager) into the office on a Wednesday afternoon. The whiteboard was half-erased from the morning IC. Claudia's printout sat between them. $682,000 per corporate FTE. And next to it, the capacity map: forty percent of the team's hours consumed by assembly.

"Walk me through what happens if that forty percent goes away."

Scott Engel (Head of Asset Management), David's boss, had joined the meeting uninvited. He had seen the printout on Claudia's desk that afternoon and asked what it was.

"Before David answers that," Scott said, "let me tell you what my team's week looks like. Three asset managers, fifteen active properties. Monday and Tuesday they pull data from property managers. Wednesday is reconciliation. Thursday is writing the narrative for whoever needs it. Friday is catch-up."

He leaned forward. "My team manages a billion-two. They could manage two billion. The constraint is the eighteen hours a week each of them spends on assembly rather than their judgment. Fifty-four hours of asset management capacity going to data reconciliation. One and a half FTEs doing nothing but bridging systems."

Marcus looked at Claudia's printout. The number landed differently now. It was Scott's team rather than abstract overhead, unable to manage the portfolio the firm was building because the infrastructure consumed the hours the portfolio required.

David started at the top: assembly, formatting, reconciliation. Where the hours lived, by line. Marcus did not interrupt.

David ran the math out loud. Twenty-two people. Forty percent of their time freed up. That is the equivalent of nine additional people doing judgment work, without a single new hire. The team that currently manages $1.2 billion in AUM could manage nearly twice that. Same headcount. Same comp. Twice the portfolio.

"At $2 billion AUM, the fee revenue roughly doubles. Twenty-two people. You are past a million a head. And the only thing that changed is what the team spends its time on."

The math deserves one honest caveat before the box gets drawn around it. Freed hours convert to managed assets only when the firm redesigns the work around them. A firm that automates the assembly and changes nothing else gets nine people's worth of output arriving at a review step still sized for the old volume, and a new bottleneck where the old one used to be. The work piles up in front of the few people authorized to approve it, and the firm mistakes motion for capacity. The platform removes the assembly. The CEO still decides where the recovered hours go, who reviews what the platform produces, and what approval looks like at twice the throughput. Chapter 9 is that redesign. The $2 billion is real, and it is earned twice: once by the platform, and once by the operating model built around it.

Marcus wrote $1M / head on the whiteboard. Drew a box around it.

He described Elena Vasquez's firm. Sixteen people. Similar AUM, similar strategy. Five times the deal-screening volume. Her team ran at fifteen percent coordination overhead instead of forty. They were the same people inside better infrastructure. And that infrastructure was the reason her firm could take on the next fund without doubling headcount.

"That is about what happens to the numerator when the denominator becomes more efficient, rather than about paying people differently."

David nodded. He already knew the answer. The distance between $682,000 per head and $1,000,000 per head was a growth question rather than an efficiency exercise: what happens when the same team manages twice the AUM because the platform absorbed the assembly work?

The platform is what grows the numerator without changing the denominator. Same team, more AUM, same comp, higher revenue per head, lower comp ratio, more durable margin. The $1M-per-head firm gets there by scaling AUM with a team whose capacity has been unlocked, rather than by squeezing more out of existing revenue.

---

The platforms above $1M per head

The benchmarks exist and are public. They are simply ignored by firms that have never thought to ask the question.

At the $100M to $500M AUM tier, firms typically run $300,000 to $600,000 per corporate head. Teams of eight to fifteen people, fee revenue of $3 million to $7.5 million. This is the scrappy phase. Every person wears three hats. The coordination overhead is enormous but tolerable because the team is small enough to compensate with brute effort.

At the $500M to $2B tier, Marcus's tier, the range widens: $500,000 to $1 million per head. Teams of twelve to twenty-five, fee revenue of $7.5 million to $30 million. The firms at the top of that range are the platform-built ones. Same AUM, fewer people, higher output per person. The firms at the bottom are the ones still running on brute effort, hiring linearly as AUM grows.

At $2B to $10B, lean firms already exceed $1 million per head. Many are closer to $2 million. The infrastructure made the scaling possible; the scaling made the number inevitable.

At the top, the $10B-plus platforms, the capital allocators running multiple strategies across asset classes, revenue per head reaches $1.4 million to $3.9 million.

Sources: PERE 100 rankings (2024), SEC 10-K filings (Blackstone, Starwood Capital, Ares Management), McKinsey Global Private Markets Review (2024), Preqin operational benchmarking data. Revenue-per-head ranges derived from total fee revenue and disclosed corporate headcount across publicly reporting firms and PERE-ranked operators.

The tiers read as a firm architecture question. Firms that spend their senior attention on the metric without touching the architecture can spend years attempting to optimize a symptom. Firms that spend their attention on the architecture soon discover that the metric has already begun to move.

---

The reclamation map

Recovered capacity has to go somewhere measurable, or the diagnostic stays theoretical. Walk it through Marcus's actual deal funnel.

Chen Capital sees roughly 380 to 450 inbound deal opportunities per year. Brokers, off-market sourcing relationships, capital partners forwarding deals, conference inbound, repeat sellers. From that sea of inbound, the team properly screens 80 to 100 each year. Properly screened means the deal got past first-look filters and a real underwriting pass. Of those 80, roughly 10 to 12 advance to LOI. Of those 10 to 12, the firm closes 6 to 8 acquisitions per year. ~400 inbound, ~80 screened properly, ~10 LOI, ~7 closes.

Where the platform changes the funnel is at the screening step. Twenty recovered analyst hours a week is a thousand hours a year; at roughly twenty hours per proper screen, that is fifty more deals screened properly, from 80 to 130 per year. Hold the firm's historical close rate against the deeper pool and the firm closes 9 to 11 per year instead of 6 to 8. Two to three additional acquisitions per year, at better terms because the team has more time to work each deal, register in returns and firm growth.

In addition, twenty additional hours per quarter of recovered IR capacity, redirected from manual brief assembly to LP relationship-building, registers in capital formation timing for the next fund. Strategic asset-management decisions made proactively rather than reactively register in business plan outcomes across the portfolio.

The platform also changes what the headcount does by freeing the capacity that lets the firm take on more. Each role description changes more often than the headcount does. The harness absorbs the work one level below each role's title; each person rises to do the work the title was supposed to describe. The analyst becomes a field operator instead of a desk operator. The asset manager moves from firefighting upstream breakdowns to managing portfolio value at strategic altitude. The IR associate stops chasing down late K-1s and starts cultivating LPs toward the deals that fit them. The founder gets back to vision, key customers, and strategic guidance of growth.

---

One scout, one workflow

Marcus walked back to his office. He did not sit down right away. He stood at the window and held both pains at once. The backward-looking one: the millions in unrealized capacity that had been sitting on the table every year, invisible because every firm he benchmarked against was paying the same tax. And the forward-looking one: the growth trajectory his firm had missed, the compounding years where the platform-built firms pulled ahead while he added bodies to match each increment of scale.

The Coordination Tax had cost him money and time. Time, in a compounding business, is the one resource you cannot recover. Until you stop paying the tax. Then the compounding starts working for you instead of against you.

He pulled Elena Vasquez's business card from the drawer where he had filed it after Scottsdale. He had been thinking about her firm's five-times deal-screening volume and five-day quarterly reports as technology stories. He saw them differently now. They were leverage stories. Same team size, same market, different relationship between people and infrastructure. Her firm was already past $1 million per head, because her platform let sixteen people manage the AUM that would have required twenty-five at his coordination overhead. Her number, multiplied across every year the platform had not existed at his firm, was the size of the debt.

The Director of Acquisitions hire he had drafted had been the visible piece of that invisible cost. He had been about to spend half a million dollars a year, fully loaded, on a candidate profile that already belonged to a different firm than the one he was building. The hire still made sense, but the role had changed.

He needed to stop paying the tax that had been compounding against him every quarter for the last decade, rather than hire his way to the next level. The platform was the end of a debt he had been silently accruing, rather than a future investment.

The harder question was how to start. A firm-wide platform mandate would produce a firm-wide platform memo and very little adoption. He had watched that movie at peer firms. The CEO declares the AI initiative, the team nods, six months later the consultant deck is sitting unopened in a shared folder. He needed proof before mandate. A live result on one workflow that the rest of the team could see, touch, and compare to the old way.

A piece he had read on AI adoption came back to him. Nate B. Jones called the model Scouts and Strike Teams.¹ The initiative starts with a single person: a scout. The scout is rarely the CTO. The scout is the person closest to the pain, curious enough to try something different, and credible enough that colleagues pay attention to the result. The scout's job is to take one broken process, rebuild it on the harness, and test the AI inside it. A quarterly variance report assembled in three hours instead of twelve days. A deal screening drawn from the firm's actual comp database. An LP brief that reflected each investor's preferences without rebuilding from memory. The result is the proof of concept: a live workflow rather than a pilot deck. That proof creates the strike team: two or three people who work on the same process, see the scout's result, and say I want that. By the time the CEO formalizes the platform as a firm-wide initiative, half the team has already adopted it. Value gets proven before the mandate.

Marcus made a short list at the window. Scott, who had rebuilt the same variance report sixteen quarters running and could describe the time cost down to the minute. Priya (investment associate), who had walked into the firm fluent in what an institutional platform looked like and had been holding the templates in her head for a year. Nathan (VP of Acquisitions), who had been guessing at the buy box for eighteen months and would recognize a codified version the moment he saw one. Three candidates. He would start with one workflow. The right scout would self-select once the workflow was named.

One scout. One workflow. One proof point that the freed hours were real.