How to Verify Seller Add-Backs Before You Close

The short answer

An add-back is real only if three things are true:

1. The money was actually spent, and a document proves it. An invoice, a bank line, a payroll record. Not a spreadsheet row.
2. It stops when you own the business. A cost that will keep happening under you is not an add-back. It's an expense.
3. The business doesn't need it to produce the same revenue. If cutting it would cost you customers, staff or equipment, it was never discretionary.

Anything that fails one of those tests is not an add-back. It's the seller's negotiating position, and you're entitled to treat it that way.

Every adjusted figure you've been shown — SDE, adjusted EBITDA, "owner benefit" — is reported earnings plus a list of add-backs. The price is a multiple of that figure, and your loan is sized to it. So every dollar of add-back that doesn't hold up is a dollar multiplied into the price and then borrowed against with a personal guarantee. A technically correct EBITDA can still be economically false.

Accepted, contested, or deal-killing

When I review a seller's add-back schedule, every line ends up in one of three buckets.

Accepted. Documented, one-time or owner-specific, and not needed to run the business. Example: the owner's personal car insurance run through the company, matched to the insurer's entries in the books.

Contested. Real money, but the seller's treatment is too generous — or the seller got it wrong in either direction. These go into the price negotiation.

Deal-killing. Undocumented, recurring, or hiding a cost the business needs. When enough earnings sit in this bucket, the business you'd be buying is not the business in the marketing package.

The four tests for every add-back

Run each add-back through these in order. Most fail on the first.

1. The source-document test

Ask: what paperwork proves this, and who issued it? A third-party document — a vendor invoice, a bank statement line, a payroll register, a court filing — beats anything the seller or broker prepared. If the only support is the add-back spreadsheet itself, the add-back is unsupported.

Make the request easy and specific. On one deal the seller couldn't produce invoices behind a big piece of revenue and suggested I fly across the country to dig through boxes. I asked for any ten invoices, photographed on a phone, within two days. They never came. When the bar is that low and the seller still can't clear it, you have your answer. (Video: He Faked $1M in EBITDA, Here's How I Caught It)

2. The one-time test

"One-time" is the most abused word in a marketing package. Line up the same expense across three years plus the trailing twelve months. A "one-time" repair that shows up every year is a repair budget. A "non-recurring" legal fee that recurs is the cost of how this business operates.

3. The owner-specific test

Owner perks are legitimate add-backs: the seller's car, phone, family on payroll who do no work, personal travel. Owner pay depends on which earnings figure the price is built on.

• If the deal is priced on a multiple of EBITDA, add back only the part of the owner's pay above what it would cost to replace him. If the owner takes $250,000 and a general manager to do his job would cost $150,000, the add-back is $100,000. The business needs that role filled whoever owns it.
• If the deal is priced on a multiple of SDE (seller's discretionary earnings), add back one owner's full compensation. SDE assumes one owner-operator runs the business, so his pay is part of what you're buying — and what you'll live on.

Check which one your price is built on before you read the add-back schedule. Mixing them — a full owner add-back with an EBITDA multiple — is one of the quickest ways to overpay. If there are two owners on payroll, only one comes back under SDE.

The same logic applies to anyone who will stay. On one deal a five-person management team was paid about $300,000 below what it would cost to replace them. The buyer assumed that didn't matter because they planned to stay. But you can't count on people working below market forever, and if one leaves you replace them at market. That gap went back to the seller in the price negotiation.

4. The business-necessity test

Some of the most expensive add-backs aren't on the schedule at all. They're costs the seller cut to make the last twelve months look better: staff, maintenance, marketing, equipment. Compare cost lines against revenue over time. If revenue held steady and a major cost fell sharply, ask what changed — and ask for proof.

What a real add-back schedule looks like

Here is the add-back review from Guardian's Sample Quality of Earnings Report, a plumbing company with about $5.4 million of revenue. Twenty-four lines were reviewed. A selection:

• Fuel. Seller claimed: Owner's personal vehicles, seller's estimate. What we found: Re-estimated on 2 cars, 15,000 miles a year, 20 MPG, $3.00 a gallon. Result: Revised: $4.5K a year.
• Auto insurance. Seller claimed: Owner's personal car insurance. What we found: Matched to the insurer's entries in the insurance account. Result: Accepted.
• Phone. Seller claimed: Personal lines on the company plan. What we found: 4 personal lines at $900 per line per year. Result: Accepted: $3.6K a year.
• Utilities. Seller claimed: Owner's home water bill. What we found: City water for the owner's home, billed to the business. Result: Accepted.
• Life insurance. Seller claimed: $40K in each of three years. What we found: Two of the three years were booked as owner distributions, not expenses — there was nothing to add back. Result: Cut to one year: $40K.
• Parking lot paving. Seller claimed: $33K. What we found: Booked to fixed assets as capex, never expensed. Result: Rejected: $0.
• Officer wages. Seller claimed: Owner's salary. What we found: The seller's own schedule missed a $50K payment booked to direct labor. Result: Increased by $50K.
• Truck purchase. Seller claimed: Not claimed. What we found: About $70K for two vehicles was expensed instead of capitalized. Result: Added: $70.1K.
• Miscellaneous. Seller claimed: $2,671. What we found: Funeral costs for the owner's mother. Result: Accepted.
• Property taxes. Seller claimed: Building taxes. What we found: Rolled into the new rent going forward. Result: Added back.
• Rent. Seller claimed: nothing. What we found: The business will pay $5K a month in rent after closing, never expensed before. Result: Deducted: $60K a year.
• Employee advances. Seller claimed: nothing. What we found: Advances on the balance sheet the company agreed to write off. Result: Deducted.

Three things to notice:

• Add-backs cut both ways. We cut the life insurance and rejected the paving, and we also found a $50K wage payment and a $70K truck purchase the seller missed. The job isn't to make the number small. It's to make it right.
• Some adjustments go the other way. Rent the business never paid but will pay under you is a cost, not an add-back. It came off earnings.
• Every line has a reason a buyer can check. If you can't trace a line back to a document or a stated assumption, you're taking it on faith.

Want the full report? Download the Sample QoE Report.

The add-backs that fail most often

Staff cuts that only look like efficiency. On one deal, trailing-twelve-month revenue was close to 2019 levels but labor costs were about $400,000 lower. The broker said the company had "become leaner." Halfway through, the seller mentioned he paid some contractors in cash "from time to time." The likely story: staff had been let go on paper and paid under the table. Cash payments can't be traced, so there was no way to know how much real labor cost was missing. If the hunch was right, the cash payments would have eaten the earnings entirely. The lender also flagged the gap, and the deal didn't go through. (War Story: Under the Table)

Current-period changes presented as history. A distressed software business had about $1 million of revenue and broke even. Then came the story: staff just cut, a new contract just signed, a merger with a sister company that would halve costs. None of it was in the historical financials. The deal team wanted $800,000 of add-backs accepted on the strength of a spreadsheet. A quality of earnings presents what happened, not what someone hopes will happen. If a change is real, let it show up in the numbers before you pay for it. (Video: $800K of Wishes, Not Add-Backs)

Everything added back. The most outrageous schedule I've seen came from a well-known brokerage: a labor recruiting company with about $1 million of revenue marketed at $400,000 of SDE. The add-back sheet added back 100% of operating expenses. The sellers' explanation was an "unorthodox tax strategy" they couldn't explain on the call. Instead of walking, the buyer and I rebuilt the P&L from the costs he expected to run the business. At that rebuilt SDE the price still worked, for him and for the lender. Fake add-backs don't always kill a deal. Sometimes they just reset the price. (War Story: Add It ALL Back)

"Revenue" that isn't revenue. Not an add-back, but the same game. A West Coast security business with $3 million of revenue claimed $1 million of profit on a $4 million price. It passed through QuickBooks, the bank statements and the tax returns. The lie was revenue that no invoice supported. Bank statements show cash in and cash out; they don't show where a deposit came from. Money from a relative or a real-estate account deposited as "sales" drops straight to the bottom line. The bank had already underwritten the loan and was pushing the buyer to close. Lender approval does not mean the business is safe. (Proof of Cash: How a Seller Faked $1M in EBITDA)

What to ask the seller for

One line per item. Ask for these as soon as the LOI is signed.

• General ledger detail for every account that carries an add-back, for each year on the schedule — exported from the accounting system, not retyped.
• Bank statements for every business account, same periods, from the bank.
• Business tax returns, as filed, for the same years.
• Payroll registers and W-2/1099 summaries, by person.
• Invoices and contracts behind each add-back over a threshold you set, plus a sample you pick.
• The add-back schedule itself, with the seller's written explanation for each line and who prepared it.
• Related-party list: any vendor, landlord, lender or employee connected to the owner.

Red flags

• No supporting document, or the document was prepared by the seller or broker.
• The payee is related to the owner.
• Large adjustments booked near year-end.
• An expense deducted on the tax return but called "personal" on the add-back schedule — or the reverse.
• "One-time" items that appear every year.
• Add-backs for things that were never expensed, like capital purchases or owner distributions.
• Cost lines that fell sharply while revenue held.
• Cash payments to staff or contractors.
• Requests that you come on-site to see records that could be scanned.
• Explanations that rely on a future event: a new contract, a merger, a planned cut.

Who can confirm an add-back

The seller's tax accountant doesn't settle it. The CPA who prepared the returns worked from the seller's numbers. Only the independent person who made the underlying entries, backed by third-party documents, can stand behind them.

The broker doesn't settle it either. The broker is paid at closing. An add-back schedule from the broker is a sales document.

The lender doesn't settle it. A lender's approval tells you the loan fits the lender's rules. It doesn't tell you the earnings are real.

If the add-backs are material: protect yourself in the purchase agreement

Diligence tells you what's real. The purchase agreement decides who pays if something you relied on turns out not to be. Talk to your M&A attorney about:

• A seller note with a right of offset, so you can reduce payments if represented earnings prove false.
• An escrow or holdback sized to the add-backs you're taking on faith.
• Representations and warranties on the financial statements, written to cover the add-back schedule.
• Post-close access to records, so you can check after you own it.

None of these replace verifying the add-backs before you sign. They're the backstop.

What to do with what you find

There are three good outcomes, and all of them are wins.

• Close with conviction. The add-backs hold up and you know it.
• Renegotiate. Contested add-backs come out of the price. "The seller actually reduced their price by $1.5M after going over the reports," said Kishon Clayborne.
• Walk. Collin Hathaway's add-on acquisition was presented at $800,000 of EBITDA. The real figure was $0, and the recommendation was not to buy.

The bad outcome is the one where you find out after closing.

FAQ

How do I check whether a seller's add-backs are real? Trace each one to a third-party document, confirm it won't recur under your ownership, and confirm the business doesn't need it to produce the same revenue. Line the expense up across three years plus the trailing twelve months to catch "one-time" costs that repeat.

Is the owner's salary an add-back? It depends on how the deal is priced. On a multiple of EBITDA, add back only the amount above what a replacement would cost. On a multiple of SDE, add back one owner's full compensation.

What add-backs are usually legitimate? Documented personal expenses run through the business — the owner's car, phone, home utilities, personal travel — family members on payroll who don't work there, and genuinely one-time costs, each backed by paperwork.

Can an add-back be something that was never expensed? No. If a cost was booked as a capital purchase or an owner distribution, it never reduced earnings, so there's nothing to add back. In our sample report, a $33K paving project and two years of life insurance failed for exactly this reason.

Does SBA or bank approval mean the add-backs are verified? No. Approval means the loan meets the lender's criteria. Lenders have underwritten deals where the earnings didn't exist.

What if the seller won't provide documents? Make the request smaller and easier, with a short deadline. If they still can't produce it, treat the add-back as rejected.

Have a signed LOI?

You get Elliott on your deal. No B-Team.

Need Diligence Now →

Related: What Does an ETA Quality of Earnings Cost? · How to Choose a Quality of Earnings Provider · War Story: Add It ALL Back · War Story: Under the Table — Reducing Staff to Boost EBITDA · War Story: Sandbagging Salaries · Why Is Adjusted EBITDA So Important in Quality of Earnings?

Videos:

• $800K of Wishes, Not Add-Backs
• He Faked $1M in EBITDA, Here's How I Caught It