Adjusted EBITDA is your earnings after a buyer removes the accounting choices that only made sense for tax season.
The biggest one is you. Your actual pay comes out. What it would cost to hire someone to do your job goes in.
That single swap moves most owners’ numbers by six figures. Sometimes in the direction they were hoping for.
Here is what it costs when you get it wrong.
Your P&L was built to lie to you
Not maliciously. On purpose.
Every year you sit down with your CPA and the goal of that meeting is to make your business look as small as possible. Defer the revenue. Accelerate the expense. Run the truck through the company. Pay yourself in a way that minimizes the check to the IRS.
That is good tax work. Your CPA is doing the job you hired them for.
But then one day you decide you want to sell. And you pull up that same P&L, find the net income line, multiply it by whatever multiple you read about on Reddit, and arrive at a number you tell your spouse over dinner.
That number is fiction. It is fiction in both directions, and most owners only find out which direction during diligence, when there is a term sheet on the table and no time left to fix anything.
Tax accounting exists to shrink the business. M&A accounting exists to isolate transferable cash flow. They are two different questions asked of the same data, and they produce two different answers.
Your books are not wrong. They are answering a question the buyer did not ask.
For the full picture of how buyers price the asset, start with MSP valuation multiples.
How private equity calculates a true MSP EBITDA multiple
A buyer is not purchasing your net income. They are purchasing the cash flow that survives your departure.
That distinction drives everything.
When a transaction accountant opens your books, they are running one test on every line: does this expense continue after the owner leaves, and does this revenue continue after the owner leaves? Anything that fails the test comes out. Anything that only exists because you own the place comes out.
The largest single item is almost always you.
Service Leadership’s Adjusted EBITDA methodology replaces actual owner pay with Fair Market Compensation for Same Position. If you pay yourself $500,000 but a hired CEO doing your actual job would cost $250,000, that gap is not profit. It is an accounting artifact of you owning the company. It gets added back.
It also runs the other way, and this is where owners get hurt. If you pay yourself $80,000 because you have been living on distributions and reinvesting everything, a buyer does not see a lean operator. They see $170,000 of real expense you have been hiding from your own P&L. That comes off. Your EBITDA drops. So does your price.
The number you have been proud of moves, sometimes by half a turn of the multiple, before a single negotiation starts.
Where the average MSP actually sits
Real numbers, so you can locate yourself.
The average Infrastructure-Managed Services firm ran 13.7% Adjusted EBITDA in 2025, up from 10.8% in 2021, per Service Leadership’s 2026 Profitability Report. Their 2026 budgeted figures break out by quartile: bottom quartile 7.5%, median 14.3%, best-in-class 24.3%.
Read those three numbers again.
Best-in-class firms earn roughly 2.5x the Adjusted EBITDA percentage of their median peers. Not 10% better. Two and a half times. And Service Leadership’s data holds that gap regardless of size, age, geography, client profile, or vendor stack.
So the story you tell yourself about your market being different, your clients being cheaper, your region being harder, is not supported by the data. Firms your size in markets like yours are earning 24.3% while you earn 14.3%. The variable is not the market.
Meanwhile, roughly a quarter of IT solution providers are losing money at any given moment.
The spread is not luck. It is operational maturity, and it shows up in the books before it shows up in the price.
The architecture of add-backs
Here is what actually survives a Quality of Earnings review.
Owner compensation. The gap between what you pay yourself and Fair Market Compensation for Same Position. This is the cleanest add-back in the deal and usually the biggest. It requires a defensible comp benchmark, not your opinion of what you are worth.
Discretionary expenses. The vehicle. The country club. The spouse on payroll who does not work in the business. The trip to Cabo that got coded as a sales conference. These come back, but only with documentation showing they are genuinely non-operational.
One-time operational friction. Litigation that resolved. A tool stack migration you will not repeat. A rebrand. A single catastrophic client remediation. The test is whether it recurs. If you have migrated your PSA three times in five years, that is not one-time. That is your operating rhythm, and a buyer will say so.
Chart of accounts structure. This is the one owners skip, and it quietly costs the most. Service Leadership’s Normalized Chart of Accounts requires all service payroll, billed and unbilled, to sit in COGS, sales and marketing broken out inside SG&A, and manager payroll allocated by function. Most MSPs have engineers scattered across COGS and SG&A depending on who was doing what that quarter.
If your accounts are not normalized, your gross margin is not real. And if your gross margin is not real, the buyer cannot benchmark you against anything, so they price you on the assumption that you are below median.
You get discounted for being unmeasurable.
For what happens when the transaction accountants arrive, read Quality of Earnings for MSP.
Why messy add-backs create an owner discount
An add-back you cannot document is worse than an add-back you never claimed.
Claim thirty of them, support twelve, and you have not lost eighteen adjustments. You have lost the buyer’s trust in all thirty. Every subsequent number you present now gets a haircut, because you have taught them that your representations do not hold up.
That is the actual mechanism of the owner discount. It is not that buyers dislike founders. It is that undocumented optimism early in diligence recalibrates every assumption downstream.
Then comes deal fatigue. Four months in, third round of document requests, your controller is drowning, you are running the business and the diligence at the same time, and the buyer revises the LOI down. Most owners sign. Not because the new number is fair, but because they are exhausted and have already told everyone they are selling.
The add-backs were never the problem. The documentation was.
Fix the books eighteen months before you need them, not eighteen days.
You do not get paid for the profit you earned. You get paid for the profit you can prove.
Auditing your true run-rate
Three moves, in order.
Normalize the chart of accounts first. Everything else is built on this. Until service payroll sits where it belongs, you are benchmarking noise. This is a bookkeeping project, not a strategy project, and you can start it this quarter.
Benchmark your owner comp honestly. Find the real market rate for the job you actually do, not the title you hold. Write it down. Support it. If you are underpaying yourself, your EBITDA is overstated and you should know that now rather than in month three of diligence.
Build the add-back file as you go. Every discretionary expense gets a note when it happens. Every one-time event gets a memo the month it occurs. Reconstructing three years of add-backs from memory is how owners end up claiming things they cannot support.
None of this raises your enterprise value on its own. What it does is stop you from losing value you already built.
The average Infra-MS firm’s approximate company value grew from $3.3mm to $8.2mm over the five years ending 2025, a 25.8% CAGR, per Service Leadership’s 2026 data. That growth is real and it is available. But it only shows up in your bank account if the books can prove the earnings underneath it.
If a buyer opened your books tomorrow, would your numbers defend themselves, or would you have to be in the room explaining them?
For the wider view on where value actually comes from, see how to increase MSP enterprise value. And if concentration is your bigger exposure, read MSP customer concentration.
FAQ
What is a normalized EBITDA adjustment for a managed service provider?
A normalized adjustment removes accounting choices that do not transfer to a new owner. The primary one replaces actual owner compensation with Fair Market Compensation for Same Position, the cost of hiring someone to do that job. Others include discretionary perks and genuinely non-recurring expenses. Service Leadership uses Adjusted EBITDA as its primary profitability metric for this reason.
What expenses cannot be added back during an MSP valuation deep-dive?
Anything recurring, anything undocumented, and anything the business needs to keep operating. A PSA migration you have run three times in five years is not one-time. Deferred maintenance is not an add-back, it is a liability. Salary for a role you must refill is not discretionary. If the expense returns next year, it stays in.
How do transaction accountants verify historical cash flow run-rates?
They tie every claimed adjustment to source documentation: bank statements, invoices, payroll records, board minutes. They test for recurrence across three years, not one. Then they build a proof-of-cash reconciliation between your reported revenue and money that actually landed. Adjustments without a paper trail get stripped, and the stripping is not negotiable.
What Adjusted EBITDA percentage should a healthy MSP target?
Service Leadership’s 2026 budgeted figures for Infrastructure-Managed Services firms show bottom quartile at 7.5%, median at 14.3%, and best-in-class at 24.3%. Best-in-class firms earn roughly 2.5x their median peers, and that gap holds regardless of size, geography, or client profile. If you are under 14%, the constraint is operational maturity, not your market.