The pregnant pause, revisited

In July I wrote about a pause.

I’d ask an Operating Partner some version of the same question that I’d asked dozens of times before: when your portfolio company beats plan, how does your firm establish that your team is the reason? Then I’d wait but would get back very little. It wasn’t evasion… these are serious, capable people who spent a decade building the thing they were now being asked to measure. The pause sounded like a profession realizing it had built an entire apparatus and never built the instrument to measure it.

I thought, “OK, I’ll go find out who’s got the instrument.” So I did.

Over the past six months I’ve had about fifty conversations on a single question: Can a private equity firm actually show what its portfolio operations team is worth? I spoke to Operating Partners and Heads of the function; Deal partners who set the budgets; Industry Research people; a platform chair; some LPs; all off the record. Some of these were formal interviews. Many were the conversations I have every week anyway, where the question got wedged in between a search update and a complaint about the state of air travel during thunderstorm season. Everyone was promised anonymity, so you’ll get seats and fund sizes here, not names.

I went looking for a North Star I could hand to my clients. (I’ve placed Operating Partners for twenty years, so I’ll admit I was hoping to find one.)

There currently isn’t one. Though kind of disappointing, that conclusion wasn’t entirely unexpected.

That said, what I didn’t expect was being wrong about the part I was most confident about. In July I wrote that the pressure to measure would be coming soon from two directions: the LPs and the regulators were about to demand that firms “Show their Work”. But I certainly did miss on that one.

On the evidence I collected, the LPs aren’t demanding much of anything. Neither are the deal partners. Neither, really, are the management companies that fund these teams.

Which turns the question inside out. It stops being “Why can’t anyone measure this?” and becomes “Why doesn’t anyone want it measured?” The answer to that one is less flattering, and it implicates my own readers more than I’d like. It also points to a potentially useful destination: so let’s dig in.

A hundred firms, a hundred definitions

The honest answers came fast, and the more senior the person, the faster they came.

A solo Operating Partner building a growth-equity function from scratch, someone who has been recruited by and seen the inside of several shops, gave it to me in one line: “Really nobody has anything is the honest answer right now.” A researcher who used to sit in a megafund’s fundraising seat told me he couldn’t get real value-creation data beyond anecdote even from the inside. The head of a fifty-person platform at a megafund described his measurement as immature enough that they’d just hired an analyst to start building it.

A co-founder of a lower-middle-market buyout firm gave me the structural version, and it’s one of the more insightful things I’d heard all year. Private equity converged on one way to do deals and never converged on a way to do operations:

An Operating Partner with ten years across two firms went further. The models differ, he argued, but the work doesn’t:

Here’s the irony I can’t shake. The PE IQ in these rooms is extraordinarily high. These are firms that walk into a portfolio company on day one and install the KPI cadence, the 100-day plan, the monthly variance review, the board pack with every number tied back to the underwrite. They force rigor on everyone they own. Then they exempt the one function whose entire job is delivering that rigor. (Irony alert: exactly what we’d never tolerate from a portfolio company.)

As one deal partner put it: “If you don’t know how to run your own company, how are you going to tell these companies how to operate?”

So nobody has a method. Fine. Young discipline, hard problem, give it time. That’s the comfortable reading… except the evidence (as you’ll see) doesn’t support it.

The firms that tried, and quit

If this were just an immature discipline, you’d expect rigor to be piling up somewhere. Early adopters. Emerging standards. A direction of travel. What I found instead is rigor being abandoned, and by the people best placed to practice it.

One lower-middle-market firm built the apparatus properly. If a deal wanted operating resources, they filed a project request with an expected return on the team’s time. Some engagements were billed to the portfolio company on time and materials, with administration and accounting behind it. That’s real discipline, more than most firms have ever tried.

They took all of it apart. The Operating Partner who lived through it:

His explanation was about capacity, not conviction: “It’s hard to assign the value creation relative to the investment of time and energy. Just, it just is.”

At another firm, where the operating team is half the headcount of the entire organization, the head of the group made it policy:

He wasn’t being cavalier, and we’ll come back to why, because his reason is the strongest argument against everything in this piece.

Where real rigor does exist, it belongs to a person, not a firm. The best individual practice I found came from a commercial Operating Partner at a middle-market firm, and she picked it up out of necessity. At a prior employer, she told me, “It was hard to get paid… for the work you did,” so she got very good at tracking her marginal contribution.

Now she tries to settle attribution with the deal team before the work starts, based on how much of the outcome she’s on the hook for. If she only delivers the recommendation, she takes maybe 5% of the credit. If she builds the systems and processes, it’s shared, roughly 50/50. If she steps in and owns the outcome herself, say as interim chief commercial officer, it’s closer to 80%. In her words: “It’s small, it’s shared, or it’s mostly me.” Agreeing up front is what keeps it from turning into “retroactive storytelling.”

It’s an excellent approach. But it’s also bespoke to her. I did find a handful of practitioners like this, people keeping a private ledger out of professional pride. However, I found no firm where that discipline survives the person walking out the door. When they leave, poof. The ledger goes with them.

This is usually the point in a piece like this where I’d hand you a maturity curve: here’s where you are, and here’s the next rung up. I can’t, because the field isn’t climbing one. If anything, two of the best-resourced operating teams I spoke with are stepping back down, and they’re doing it on purpose.

What firms say instead

With no pure method, when it is asked, PE generally improvises. Ask a firm how it shows its operating team is worth the money and you’ll get some mix of four answers.

1. Case studies and anecdotes. The default everywhere. One head of function summed it up bluntly: “You’re full of case studies. You can’t do full attribution.” An Operating Partner described the LP version as “Really just show and tell.”
2. NPS scores. Survey the CEOs, the deal teams, the chairs. It’s a real customer metric. It tells you whether people liked working with you, not what you returned.
3. The fund scoreboard. How have returns been? Several heads of function treat this as doctrine: the team is part of the firm, the firm is judged on its returns, end of conversation. It’s honest. It also attributes nothing.
4. Pull. The CEOs keep asking for us. Demand is the proof.

That fourth one is the field’s safety net. When every other argument runs out, everyone reaches for it. A megafund head of function told me the firm’s co-founder agreed that the better CEOs use the team the most. The logic feels airtight: good operators seek out good help.

Then an Operating Partner with outcome data across a portfolio of roughly thirty companies told me the opposite:

His explanation was a simple four-box. Companies beating plan aren’t receptive, because they believe they’re doing fine. Companies badly missing plan can’t absorb help. When you’re drowning, you aren’t thinking about long-term strategy. The yield concentrates in the middle. His favorite example was a management team that turned down help outright, essentially telling him: “We know what the hell we’re doing. And we don’t want to be distracted… We’re going to deliver our numbers and I’ll let you know if I need help.” They delivered.

A head of function at another firm reached the same place by design rather than by observation: “If I have to stay engaged for there to be a return, that’s a broken model… their best possible answer is the management team’s got the ball and I don’t even need it.” On that account, success and involvement are inversely related by construction, because the goal of the work is to become unnecessary.

I don’t think that settles it. Pull may measure a CEO’s humility rather than their quality. Firms may simply send operating resources where the trouble is, which would make involvement a marker of a struggling asset. But neither reading rescues pull as a scorecard. At best it’s uninformative. At worst it points the wrong way. And it’s the proxy the whole industry leans on hardest.

Nobody is asking

This is the part that reframed the whole study for me. I’d assumed the attribution gap was a supply problem: the market wanted evidence and the function hadn’t produced it yet. The evidence says the opposite. Four audiences could demand this. Three of them don’t.

The LPs

The LPs are not asking, at least not in the way the industry tells itself they are. A growth-equity GP who had just closed a fund described the LP posture as wanting to hear that you have “table stakes full-time Operating Partner” capability. A senior leader at one of the largest platforms in the world, who has sat through more fundraises than almost anyone, reached the same phrase independently two weeks later:

A head of function who is in front of investors constantly: “I’m always front and center with the damn LPs every time. And they don’t press that hard on this topic, for what it’s worth.” The market data says the same thing. From a published Q3 firm census, 94% of PE firms have at least one portfolio-facing professional, and 100% of those above $25 billion of AUM do. Universal existence. Unexamined efficacy. That’s exactly what a box-checking equilibrium looks like from the outside.

There’s a simple reason for it, and nobody had put it to me this way before this year. The function is paid out of the management fee. As one Operating Partner put it: “It’s funded by the management company. So (the LPs are) sort of not paying for it, quote unquote.” Nobody audits what they don’t buy.

The exception proves the rule. The one head of function who told me his LPs really dig in also told me why: “A lot of it is self-interested. Hey, you know, I’m actually doubled down on this company. What the fuck are you guys doing to drive returns?” When an LP co-invests directly in a portfolio company, the operating team’s productivity suddenly becomes very interesting. The LP paying at the company level is the LP who asks. (He added that the larger ones are also quietly working out how to build the same capability themselves. Coopetition, he called it.)

The Deal Partners

This one surprised me most. I put the question to multiple Deal Partners, from a roughly $500 million growth fund up to a megafund. Every one of them declined the exercise. Not evasively. They just didn’t think it could be done, and each had a different reason.

• Too many unknown inputs. “You’re not going to run the experiment and not have your operating group talk to some company, right? So it’s hard to know then what would have happened.”
• Asset selection matters more. “You can’t just say, hey, we did all these projects, created all this value in this portfolio company, and not think, well, what about the 3,000 deals we saw that led to us buying that one?”
• The assumptions are unsalvageable. Pulling apart the ops group versus management versus the market and luck involves “so much nonsense in the assumptions you’re going to make as you do that, that like it isn’t even worth spending time on.”
• Macro, the CEO and culture swamp everything else. When I suggested to the senior-most deal voice in the study that, with all three in play, any measurement may be illusory to begin with, the answer was immediate: “Yeah, exactly.”

I spent months trying to explain the disagreement about attribution by fund size, seniority, culture and sophistication. It sorts on none of those. It sorts on which side of the table you sit. The people who produce attribution numbers believe in them far more than the people who receive them.

The firm itself

Then there’s the finding that should end the argument that better measurement is the answer. Proof doesn’t buy investment.

The head of one megafund operating group can document roughly a billion dollars a year of value, and nobody internally disputes it. “I can prove, I can stack it up, I can show the dollars… and no one here debates that.” He’s had two consecutive years of refused budget requests. He laid out a procurement opportunity of more than $300 million across the portfolio for his deal teams. The response, in his words: “literally crickets, just crickets.” A head of function at a different megafund produced a comparable number and was told, in substance, that the firm didn’t much care.

If better measurement were the answer, these two would run the best-resourced teams in the industry. They don’t.

A head of portfolio operations who has sat in the seat at two firms told me why he thinks that is. His previous firm measured everything. His current one measures almost nothing and invests on conviction. His read on the first one: “They were measuring it because they doubted it.” Measurement, in other words, may be a symptom of doubt rather than a cure for it. The firms that believe in the function don’t bother. The firms that don’t believe in it won’t be persuaded by a spreadsheet.

The only audience consistently asking for more rigor is the portfolio companies themselves. One firm’s own portfolio survey, with 68 responses and 84% coverage, came back with a detractor theme asking for “more rigorous value-capture (working capital, realized savings, EBITDA contribution) post-initiative.” The customer wants the number. The buyer and the allocator don’t.

The measurable half

There’s a clean answer buried in all this to a question I didn’t set out to ask. Attribution isn’t equally hard across the function. It tracks to the kind of work.

The provable value sits in the unsexy stuff. A Deal Partner at a lower-middle-market firm told me his portfolio spends $50 million a year on insurance across benefits and P&C, “And we don’t have a single person in the whole fucking ecosystem that knows anything about buying insurance.” So he hired a head of risk management. Then a general counsel to take on legal spend. That’s work with a named cost line and a before-and-after. The number falls out as a by-product.

A head of function at a growth-oriented software firm discovered the same thing, a little to his own surprise. He’d brought in a contract operator to squeeze vendor contracts, the least glamorous assignment on the board:

Compare that with the generalist seat. The same deal partner described a senior portfolio-ops hire, paid “$600,000 a year plus equity,” who missed “a bunch of really fucking basic shit” at a portfolio company. Generalist judgment work is unbounded by design. It’s the advice in the boardroom, the CEO call at 10pm, the pattern recognition that steers a company away from a mistake it never makes. No measurement method captures that, and pretending one does is how you end up with none of it believed.

The P&L or nothing

Even measurable value has to land in the right place. An Operating Partner who built a function from zero stated the bar plainly: “You don’t get credit for cost avoidance, right? You get credit for like actually improving the P&L.”

I’d go further. Nobody gets credit for cost avoidance in year two of a program. In year one it’s a win. By year two the avoided cost is simply the new baseline, and the people who negotiated it have become invisible. A senior leader at one large platform, asked about the firm’s headline value number, didn’t dispute the cost-avoidance piece. The response simply weighed it: “And that’s great, but, you know, it’s not that significant.” It wasn’t challenged. It was discounted, which is worse, because there’s nothing to argue with.

What this does to team construction

Follow the logic and it reshapes the org chart. If the measurable work is functional, bounded and tied to a cost line, then pressure for measurement pushes firms toward functional specialists: procurement, pricing, insurance, legal, data. The hiring market is already moving that way. In that same Q3 census, 83% of AI-titled operating hires came from technical backgrounds, and only 17% from the consulting generalist pool that has fed these seats for twenty years.

But that’s not the whole market, and I see the other half every week. A large share of firms still want the pod model: a vertical deal team with an embedded generalist Operating Partner who knows that sector cold. Those people are, by definition, nearly impossible to measure on anything except fund performance and aggregate portfolio EBITDA in the vertical. The firms hiring them know that, and hire them anyway, because the judgment is the point.

So here’s an uncomfortable possibility. Some firms may be solving the attribution problem not by measuring better, but by buying less of the half that can’t be measured. That’s a perfectly rational response. Whether it’s the right one is a different question.

The case against measuring

I promised to come back to the head of function who banned attribution talk in his own team. His reason deserves a fair hearing, because it’s the strongest objection to this entire piece. Measuring the work may make the work worse.

He wasn’t protecting anyone’s comfort. He was reacting to what the scorecard did to his people: “The mental gymnastics associated with and the unhealthy behavior, frankly, associated with somebody saying, I need to go make an impact here that I can go defend on the back end.” Nobody was misbehaving. They were responding rationally to the incentive. Defensible work crowded out important work.

The head of a megafund platform had warned me of the same thing months earlier, in his own idiom: “What happens when we try to measure shit is that you don’t work on important stuff.”

The clearest version came from the global head of portfolio operations at a large platform, and it’s about AI, which is what makes it so sharp. He pointed out that a $3 million distribution center, or a $20 million five-year ERP consolidation, clears a boardroom on a cost-of-capital argument. Nobody asks for a payback date. “But as soon as I say, yeah, I want to deploy an AI tool, they’re like, great, is the payback next week? Like, where, where’s the money?” Then he explained what that standard produces:

That’s a real cost and I’m not going to pretend otherwise. A demand for proof, applied unevenly, selects for small, defensible, unambitious work. And the proof standard being applied to operating teams, and to AI, is one that ordinary capital projects in the same boardrooms have never had to meet. That asymmetry isn’t analysis. It’s a rhetorical device, and Operating Partners are right to resent it.

So: no method, a field retreating from the attempt, nobody asking, proof that ultimately doesn’t fund anything, and a credible argument that trying makes things worse. A reasonable person could stop here and call it a draw.

I almost did. Then one conversation that wasn’t about attribution at all changed how I read everything above.

The shield

Put the pieces together and you should be a little confused. The LPs don’t require it. The Deal Partners reject it. The firm won’t fund on it even when it exists. The one audience that wants it has no authority. Why would anyone build it?

The obvious answer is that there’s no reward for building it. True, but not the whole answer.

The rest came from an Operating Partner who had just turned forty, ten years into the seat across two firms, thinking hard about whether to take a third. We weren’t talking about measurement. He was explaining what compensates for the economics of the job, which, as we’ll see, aren’t always great. And he said this, unprompted:

I pushed on it. Deal Partners get hung with their results. Their track record is a list of outcomes with their name on each one. Operating Partners have something closer to a shield against complete attribution for a busted deal. He didn’t hesitate: “Yeah, that’s also very true. Yeah, yeah, I would agree with that.”

Now put that next to a line I’ve been using all year, ever since a conversation about Operating Partner comp with a go-to-market veteran of the conference circuit: “As soon as your upside is known, your value is marked.”

Hold those two together and the attribution vacuum stops looking like a failure. It looks like insurance. It caps your upside and it floors your downside, and almost nobody in the seat has an unambiguous interest in collapsing it.

That’s the uncomfortable center of this study. An Operating Partner who solves attribution gives up the ambiguity that protects him on the deals that go wrong. In exchange, he gets a claim on credit that, on everything above, will be discounted, dismissed or moved past anyway. Stated plainly, it’s a bad trade, and rational people decline bad trades. The instrument I went looking for doesn’t exist not because the field is young or the math is hard. It doesn’t exist because nobody in the room is served by its existence. Including us.

To be clear, I don’t think anyone sat down and decided this. Incentives don’t have to be conscious to be real. And this rests on a small number of voices. Treat it as a proposition, not a verdict. But it’s the only explanation I found for why the vacuum persists among people with the intelligence and resources to close it.

The terms of the bargain

The shield isn’t free, and the Operating Partners I spoke with were precise about its price. Several were angry about it.

• Second-class economics. An Operating Partner late in his career, the most senior person at his firm: “We are second-class citizens in terms of comp, carry… the deal guys want to do the stuff in the beginning and they want to do the stuff at the end and they don’t really care what happens in the middle.”
• A ceiling that’s structural, not personal. The forty-year-old again, on whether the seat makes anyone truly wealthy: “It’s a role thing. It’s not going to happen.”
• Carry on a model nobody believes anymore. “The partners of the firm aren’t getting rich off carry anymore. They’re getting rich off the 2% of management fees… the Operating Partner just gets screwed because it’s just, it’s more checks, crappier investments and more work for me.”

So the bargain looks like this. You accept a fraction of the deal side’s economics for work that’s at least as hard, a ceiling you can’t negotiate past, and carry on assets you privately doubt. In return, you get a job that’s intellectually rich, genuinely useful, and insulated from the full consequences of failure in a way no Deal Partner’s job is.

You can’t hold both halves of that at once: that the function is undervalued, under-compensated and structurally capped, and that its contribution can’t and shouldn’t be isolated. “The standard being applied to me is unfair” and “I decline to be measured” aren’t the same sentence. The first is true. The second is the position the profession is actually holding, and the conflict is being resolved, right now, in favor of the people who set the carry. The shield and the ceiling aren’t two problems. They’re one object, seen from two sides.

Underwrite, don’t attribute

Here’s the good news, and it genuinely surprised me. I never asked a Deal Partner what instrument they wished existed. Three of them told me anyway, independently, at three firms and three fund sizes. Not one asked for a retrospective scorecard. Every one described the same thing: a priced scope, agreed up front.

A growth-equity GP: “Maybe we should be more, frankly, a little bit more prescriptive on, hey, We think this is a $100,000 problem, right?… if we spend $100,000, it could generate a million. So why wouldn’t we spend $100,000, right?”

An investment partner at a family-office-backed firm, describing the consulting pitch his operating teams never match: “I think importantly, it’s on the front end. It’s not backwards looking.” And the model he admires most: “Here is the scope of work, it’s on paper, are you agreeing? You can say no.”

The senior-most deal voice in the study, on what should have happened all along: “My view is always we should go to a more prescriptive model.”

Then the part that landed it for me. An Operating Partner arrived at the identical instrument from the opposite side of the table. He wasn’t asking how to measure the work. He was asking what it would take to be paid for it:

Buyer and seller, separately and unprompted, described the same instrument. We’ve spent a decade thinking that we should be auditing the past for people who are trying to price the future. That’s the whole miss.

The pieces already exist, scattered across the firms I spoke with. Nobody has assembled all of them:

• A written scope at entry, per engagement, naming the outcome and a number.
• The portfolio company can decline it. That’s what makes it honest rather than imposed.
• Attribution share agreed in advance, scaled to how much of the outcome the team owns. The commercial Operating Partner with the private ledger already does this.
• Deal-team sign-off at open and at close, so the number is agreed by the audience that will judge it.
• Optionally, fee-at-risk. Consultants already do it: “We think we can save you $6 million, and if you want to, we’ll go at risk on that.”

A scope agreed in advance doesn’t solve the counterfactual. It dissolves it. You don’t need a control group, because the baseline is the contract.

It also answers the corrosion objection, mostly. Pick the work by what’s easiest to defend afterward and you get automated accounts payable. Agree the number before the work starts, including on the ambitious projects, and the uncertainty is priced into the deal rather than litigated after it. One of those is a contract. The other is a trial. (I’ll concede the risk doesn’t disappear. A lazy version of this just moves the timid project selection to the front end. The discipline is in agreeing to big scopes, not just safe ones.)

The one control arm that does exist

There’s also one form of evidence the field already has and nobody collects. Every source conceded the control arm can’t be run. You can’t buy the same company twice and send the ops team to only one copy. But the market has quietly run that experiment anyway, through people: CEOs who have run companies under more than one sponsor. Same operator, same temperament, often the same industry, different operating groups.

The Deal Partner who pointed this out to me said his own CEOs volunteer the comparison unprompted:

That’s the closest thing to an unimpeachable signal in this whole debate. The respondent has no stake in anyone’s fundraise. It never asks anyone to carve up a number. It asks a repeat customer to rank suppliers. It won’t price anything, and it ranks rather than measures. But it’s cheap, it’s available now, and it’s the only instrument I’ve found that could tell us whether pull points the right way.

Where the reckoning actually comes from

If it isn’t the LPs, what forces the issue? Not the regulator, at least not yet. My best answer came from a Deal Partner, and a head of function at a megafund volunteered the same answer separately: the carry pool.

As Operating Partner carry closes in on deal carry, somebody is going to have to argue for it. The Deal Partner’s words: “That will be the forcing function.” The head of function: “The bigger question is probably much more around the carry allocation than the cash allocation.”

When that conversation happens, the Operating Partner who walks in with case studies and an NPS score loses it to the colleague who walks in with a number. Or worse, to a consulting quote. As the Deal Partner put it, on a $500 million deal with 20% in the carry pool: “You wouldn’t pay any consultant $100 million for the best work ever, right?… you wouldn’t pay them a quarter of that.” Perfect attribution doesn’t win that argument. It just moves it, from “did you create value?” to “could it have been rented cheaper?” A priced scope agreed up front is the only answer to that question I’ve found.

So… Show Your Work

I’m not neutral here and I should say so again. My business depends on this function being valued, which means my interests run with the people I’m challenging. That’s exactly why I think it should be said from inside the tent rather than by a consultant selling a framework.

One Deal Partner described the moment better than I can. Private equity spent thirty years importing things that obviously worked and never measuring them, because they obviously worked. That era is over:

So here’s my ask, and it’s a small one. Next time you take on a piece of work, write down what it’s worth before you start, and show it to the Deal Partner. If you’re right, you have your number. If you’re wrong, you find out early, which is worth something too. And if you want a vote at IC, want it on the terms that come with it: attributable on the deals that fail, not just the ones that work.

The vacuum has been comfortable. It’s also been expensive, and the bill is arriving in the carry conversations of the next two funds. Of course, it’s never too late to change. But the people who write it down first are going to set the terms for everyone else.

---

Method: this piece draws on about fifty conversations held between April and September 2026, a substantial share of them recorded and formally coded, plus two industry datasets. Participants included Operating Partners, heads of portfolio operations, Deal Partners, a platform-level senior leader, industry researchers and a small number of LPs speaking off the record, across firms from roughly $500 million to well over $100 billion of AUM. All participants were promised anonymity. Quotations are attributed by seat and firm type only.