There is a phrase doing the rounds in operating partner circles right now. Operational alpha. It is the new headline of every value creation deck, the answer to every LP question about where returns will come from in a flat market.
The instinct is right. The phrase is doing a lot of work it has not earned.
Operational alpha is not a slogan you put on a slide. It is the gap between what a portfolio company could be doing and what it actually does, measured and closed lever by lever. And you cannot close a gap you cannot see. Most portfolio companies cannot see their own operations cleanly enough to know where the gap even is.
I wrote earlier this year that 12 is the new 5. The math that used to deliver returns through leverage and multiple arbitrage is broken, and the required EBITDA growth has roughly doubled. That post was about why financial engineering is dead. This one is the bookend. It is about what replaces it, and why the replacement is a measurement problem before it is anything else.
The market has already conceded the point
This is no longer a contrarian argument. The industry has already moved.
Around 72% of GPs now rank operational improvement as the top value-creation lever, according to S&P Global. That is a remarkable number for an industry that spent two decades treating operations as the thing you did after the capital structure was optimized. The center of gravity has shifted.
The reason is not philosophical. It is arithmetic. Roughly 80% of GPs expect multiples to stay flat, per Bain. When the exit multiple matches the entry multiple, you cannot buy your way to a return. When debt is expensive and leverage contribution is down, the capital structure cannot carry it either. What is left is the operation. The actual business getting actually better.
So the whole industry agrees. Operational improvement is the return now. The agreement is the easy part. Almost nobody is honest about what operational improvement actually requires underneath the slogan.
Operational alpha is the sum of measurable levers
Strip the term down and operational alpha is a stack of specific levers. None of them is abstract. Every one of them is a number that has to move.
Pricing. The single highest-leverage move in most mid-market businesses, and the one most often left on the table. A few points of price discipline flows almost entirely to EBITDA.
Procurement and cost of goods. Spend visibility across categories and suppliers, then renegotiation and consolidation against that visibility.
Go-to-market efficiency. Cost to acquire a customer, sales cycle length, win rates by segment, pipeline that reconciles to booked revenue.
Customer retention and net revenue retention. The cheapest growth a company owns, and the lever that compounds hardest over a hold.
Working capital. Cash tied up in receivables, payables, and inventory that could be funding the value creation plan instead.
Each of these is a lever you can name, baseline, target, and track. That is what makes operational alpha real rather than rhetorical. It is the disciplined pursuit of a handful of numbers, each moving in the right direction, compounding across the hold.
Now ask the uncomfortable question. For each of those levers, can the portfolio company produce the number cleanly, today, the same way twice?
For most, the answer is no. And that is where operational alpha quietly dies.
You cannot improve what you cannot see
Here is the part the decks skip.
You cannot optimize pricing without accurate cost data by customer, by product, by channel. If you do not know your true cost to serve, every pricing decision is a guess dressed up as a strategy.
You cannot fix procurement without spend visibility across the whole supplier base. If the same vendor shows up under four spellings in three systems, you cannot even see the spend, let alone consolidate it.
You cannot improve go-to-market when the CRM and the finance system disagree about what closed. The pipeline number in the board deck and the revenue number in the actuals come from different places and never quite reconcile, so the go-to-market conversation becomes an argument about whose number is right.
You cannot manage retention without cohort analysis, and you cannot run cohort analysis on a customer table where the same account exists three times under three IDs.
The pattern is the same every time. The operational lever is real. The intent is real. The data underneath it cannot support the decision, so the lever never actually moves. The initiative gets reported as in progress for four quarters and then quietly drops off the deck.
Operational alpha lives or dies on whether the company can see itself. Most cannot.
The number that does not exist yet
Walk into a portfolio company two years into a hold and ask a simple question. What is gross margin by product line, this quarter, right now.
In the businesses that are going to generate operational alpha, you get an answer in minutes and it is the same answer whether you ask the CFO or the head of sales.
In the businesses that are not, you get a different sequence. A pause. A promise to pull it together. Two people and a few days of work. And when the number comes back, finance has one version and operations has another, and the meeting becomes a debate about the data instead of a decision about the business.
That second company does not have an operational alpha problem yet. It has a visibility problem that will become an operational alpha problem the moment someone tries to actually move a lever. I wrote about how to surface this in diagnosing a stalled value creation plan. The stall almost never traces back to a bad thesis or a bad team. It traces back to a company flying the plan on instruments it cannot read.
When you cannot read the instruments, you cannot tell which lever is working. So you guess. And then you reorganize around the guess and call it value creation.
Why this gets harder, not easier, with AI
The current enthusiasm makes the problem worse before it makes it better.
Every operating partner wants AI in the value creation plan. Pricing models, demand forecasting, churn prediction, automated reporting. All of it sits directly on the same operational data we have been talking about. AI does not relax the data requirement. It raises it.
A pricing model trained on cost data that is wrong produces confident, precise, wrong prices. A churn model built on a customer table with triplicate records learns from noise. The model amplifies whatever is underneath it. Clean foundations get sharper. Dirty foundations get expensive new ways to be wrong faster.
So the firms chasing operational alpha through AI and the firms chasing it through classic operating levers end up needing exactly the same thing first. A company that can produce its own numbers cleanly, consistently, and on demand. There is no version of operational alpha, traditional or AI-enabled, that skips this step.
Making operational alpha real
The fix is not a twelve-month data transformation. It is a discipline, and it runs in weeks, not years.
Pick the five to seven metrics the value creation plan actually depends on. Not the forty on a dashboard. The handful that, if they move, mean the thesis is working. For most mid-market businesses that is revenue by segment, gross margin by product or service line, net revenue retention, sales cycle or pipeline conversion, and one or two drivers specific to the business.
For each metric, establish a single source and a single definition that finance, sales, and operations all sign. The goal is not perfect data. It is one number per metric that nobody relitigates in the room. Most of the value here comes from killing the second and third versions of the truth, not from buying new tooling.
Then connect every operational lever to one of those numbers, with a baseline and a target. Pricing connects to gross margin. Go-to-market connects to sales cycle and segment revenue. Retention connects to net revenue retention. The levers that cannot be connected to a number are the ones to question first, because they were never measurable to begin with.
That is the whole discipline. A small set of trusted numbers, every lever wired to one of them, reviewed on a cadence. It is unglamorous and it is the difference between operational alpha as a result and operational alpha as a word on a slide. If you want to see how ready your portfolio is to run this way, the VCP Data Score measures exactly this gap.
The bookend
A year ago the case to make was that financial engineering was finished. That case is settled now. The numbers in the mid-market PE figures that matter made it for me, and the 72% of GPs naming operations as the top lever made it for everyone else.
The case to make now is harder and less popular. Operational alpha is real, and almost everyone chasing it is chasing it blind.
Returns come from operational improvement. Operational improvement comes from moving specific levers. Moving a lever requires seeing the number underneath it. And most portfolio companies cannot see their own numbers cleanly enough to know which lever to pull.
The firms that win the next cycle are not the ones with the best operational alpha slide. They are the ones whose portfolio companies can answer, in minutes and without an argument, what their margin is by product line this quarter. That sounds like a low bar. In practice it is the bar that separates the funds that talk about operational alpha from the funds that book it.
This is the same discipline behind a data-driven value creation plan. Build the number cleanly first. Then the levers have something to stand on.