EBITDA lift is the value a private equity owner adds to a portfolio company's earnings during the hold, above what the business would have earned on its own. In an operations-heavy portco it comes from two places: earning more per unit sold, and spending less to produce and deliver each unit. Everything else is financing and multiple arbitrage, which you do not control once the deal closes.

If you are the operating partner, you own this number. The CEO reports it, the CFO defends it, the COO delivers it, but you are the one who wrote it into the investment committee memo, and you are the one who has to explain the variance at the annual meeting. So this page is written first for you, and then for the committee you have to move.

Here is the thesis, stated plainly so you can argue with it: in operations-heavy mid-market companies, the durable EBITDA lift is not in another cost-out program and not in the model you saw in a demo. It is in the long manual chains your most junior people run every day. Those chains never make it into the value-creation plan because they are invisible from the boardroom, and that is exactly why they survive year after year.

Which lever actually moves the number in an operations-heavy portco

The published analyses disagree on the size of every lever, and the dollar figures are noise. Pricing advocates say pricing has more leverage than cost. Operations advocates say throughput compounds with longer holds. ESG advocates say materiality assessments deliver a lift. They are all describing the same tree from different branches.

Strip the noise and one thing is true across almost every operations-heavy business we have opened up: the largest recoverable margin is trapped in process, not in the P&L line you are staring at. The cost is not the software license or the headcount number. It is the rework, the six-figure mistake a junior person makes on a checkbox in step 34 of a 50-step process, the exception that takes three people two days to clear, the customer who churned because the quote came back late.

The reason that margin never shows up in a cost-out plan is that it is not a line item. It is spread across a chain of manual steps where no single step is expensive but the chain is. A useful diagnostic on any process in the building: count the steps, count the people who touch it, and ask what the worst single error in that chain has cost you in the last twelve months. If the answer is six figures and the person running that step is the newest hire, you have found EBITDA that no spreadsheet will surface.

The exit math is the only reason this matters as much as it does. The lift is multiplied at exit by whatever multiple you paid, so a modest, durable operating gain compounds into real valuation. That is arithmetic, not a forecast. It is also why a lift that reverses the day you sell is worth almost nothing. A buyer's diligence team will find a lift that only exists because you were personally holding it together, and they will not pay for it.

Why the value-creation plan stalls at the handoff, not the thesis

Most EBITDA-lift plans are not wrong. They stall.

The classic pattern: you hire a strong ops lead or bring in a consultancy. They map the operation, and the map is excellent. They restructure, they find the real problem, they build a plan. Then they leave before the fix ships. The map goes in a drawer, the process reverts to how the junior team already knew how to run it, and next year's operating partner commissions the same map again. You have paid for diagnosis three times and delivery zero times.

The AI version of this failure is worse because it looks like progress. Someone runs a pilot. The demo works. Everyone in the room is impressed. Then the pilot hits the line marked "production" and nobody owns the last mile: the exception handling, the integration into the real system of record, the judgment call when the agent is 88 percent sure instead of 100 percent. The pilot stalls there, and now you have a line item in the value-creation plan that reads "AI initiative" with nothing running behind it.

Both failures share one root cause. Nobody owned the judgment. The tool, the model, the map, all of it is roughly 5 percent of the gap. The other 95 percent is someone senior deciding what the process should do when reality does not match the happy path, and staying long enough to make that decision hold in the live operation.

So the decision rule for any lift initiative is not "is the technology good." It is "who owns this in month six, and are they still here when it breaks." If you cannot name that person, the initiative is a diagnosis, not a lift.

What a real operational lift looks like when it ships

The shape of the work is consistent. Take a long manual chain, move the hands-on-keyboard steps to agents, and keep your experts on the judgment calls where a wrong answer is expensive. You are not replacing the expert. You are removing the 40 steps of copying, checking, and re-keying that stand between the expert and the decision.

A few patterns from work we have delivered, so this is not theory. In a PE-backed demand-generation company, a 50-step quote-to-cash process collapsed onto an agent-run spine, which removed 12,450 manual sourcing events a year. In a PE-owned automotive-software platform, data coverage on integration designs moved from 53 percent to 81 percent, which is the difference between a process that needs constant human patching and one that mostly runs. At a research firm, a document that used to take an analyst most of a day became a client-ready deck in 25 minutes at over 90 percent accuracy.

Notice what those have in common. None of them are cost-out through layoffs. Each one is capacity created inside a chain the business already ran, which is why the lift holds after the person who built it steps back. That is the test of a durable lift: it survives a diligence team that assumes every good number is being held together by hand.

What good looks like, concretely: the chain is documented because an agent runs it, the exception rate is measured because the system logs it, and the expert's time has moved from re-keying to deciding. When a buyer's diligence team asks how the process works, the answer is a running system, not a person's memory.

Is this predictable, and who carries the risk

This is the CFO's question, and it is the right one. An EBITDA-lift initiative that cannot be priced and cannot be dated is a research project, and research projects are where scope creep and overrun live.

Predictability is possible, but only under a specific operating discipline. Every change to a live process should run first in a copy of the real environment on real data, and every check the process depends on should replay before anything ships to production. That is what makes a fixed price honest rather than a sales number: you are not guessing what will break, you are proving it against the actual data before it touches the operation. If a vendor offers you a fixed price without that discipline underneath, the fixed price is a story.

The risk question is really three questions. The COO wants to know it will run and not disrupt the operation, which is answered by the copy-of-production discipline above. The CTO wants to know about integration cost and lock-in, which is answered by who owns the system at the end: if you do not own what shipped, you have bought a dependency, not a lift. And the operating partner wants to know it will not turn into open-ended consulting fees, which is answered by anchoring the engagement to a named KPI with an accountability that survives a miss, not to hours billed.

How Salfati Group would approach this as a Mandate

We would scope one operations-heavy chain where the lift is real and the failure mode is expensive, and take it on as a fixed-price, fixed-scope Mandate anchored to a KPI you would defend in a board meeting. A named architect owns it end to end, agents do the work, and your team owns the system when it ships. The Outcome SLA is simple: if we miss the target, we keep working at no additional cost until the named KPI ships. The date is set with the scope, not promised in a range. If it clears, the Vigilance Layer keeps it running and surfaces the next chain worth the same treatment.

If you have a value-creation plan with an operational line item that has been diagnosed twice and shipped zero times, apply for a Discovery call and we will tell you whether it is a Mandate or not.

Sources

1. 1. Private Equity Value Creation Starts on the Shop Floor- **The exit multiple amplifies every gain.** A $5M EBITDA lift at 7x is $35M in valuation, so modest operational wins compound into real exit value. ... Private equity value creation comes down to two levers: revenue growth and EBITDA improvement. ... Most industrial businesses are valued as a mult...
2. 2. New EBITDA Playbook for PE Manufacturing, Retail & CPGWhile these levers often reinforce one another, each addresses a distinct source of value: Cost structure, decision quality, operating scalability, revenue quality and speed of realization. ... For most portfolio companies, the earliest and most reliable EBITDA uplift still comes from structural eff...
3. 3. How PE Firms Improve EBITDA | Guru Startups Market Intelligence ...Private equity and venture-backed platforms routinely pursue EBITDA uplift as the central axis of value creation, selecting levers that harmonize top-line growth with operating leverage. In practice, the most durable EBITDA improvements emerge from a disciplined blend of revenue expansion, margin re...
4. 4. Architecting Double-Digit EBITDA Growth in Private ...In a macroeconomic environment characterized by 8–9% borrowing costs and reduced leverage capacity, the historical requirement of 5% annual EBITDA growth to achieve a 2.5x Multiple on Invested Capital (MOIC) over five years is obsolete. Today, firms must double that growth rate, targeting a 10–12% E...
5. 5. How Can Private Equity Firms Increase Their Portfolio Value ...Analysis across thousands of Private Equity investments shows that **the EBITDA leverage of an improvement in pricing typically impacts EBITDA by 10-12X**. In comparison, a reduction in fixed or variable costs for the average PE investment has 2-5X leverage on EBITDA. When you achieve a 2% effective...
6. 6. Increase EBITDA by optimizing portfolio company operationsGiven today’s longer hold periods, many PE firms are finding a more effective approach to increasing EBITDA: strengthening operational performance and, more specifically, increasing production throughput. ... With longer holds, strengthening operations performance by increasing throughput can offer ...
7. 7. Increasing EBITDA by Optimizing IT costs for Private Equity ...To summarize, implementing strategies to reduce IT costs can significantly enhance EBITDA rates and levels in the context of PE divestment. The previous report provides quantitative findings on the implementation of IT cost reduction strategies in three companies, demonstrating that all of them lead...
8. 8. ESG Materiality Assessments: How PE Firms Target 4-7% EBITDA LiftIn this analysis, we examine how researchers found that ESG initiatives targeted via materiality assessments appear to lift EBITDA by 4-7% through cost reductions and revenue growth in areas like energy efficiency and employee well-being. ... They're delivering a 4 to 7% lift in EBITDA. ... {ts:205}...

Reviewed by David Fialho·25 August 2026