Normalized EBITDA — the add-back battleground.

The purchase price of an agency almost always reduces to a single number multiplied by a market multiple. Get that number wrong and a three-times multiple turns a $50,000 error into $150,000 of overpayment. So the highest-leverage hour of financial diligence is the one you spend rebuilding the seller's earnings from the documents that prove them — tax returns, payroll registers, bank statements, and the agency-management reports behind them — rather than from the summary the seller hands you.

This piece walks the rebuild: the distinction sellers blur, the test that anchors every add-back negotiation, the three levels of scrutiny that tell you how hard to push, and the two ratios that catch a number that has been quietly engineered.

§ 01 · The number sellers quote vs the number you buyDiscretionary earnings is not your earnings.

Sellers like to talk about Seller's Discretionary Earnings because it is a large, friendly figure — operational profit blended with the owner's salary, bonuses, and perks. It answers the question "what does the current owner take home?" That is the wrong question. The one that decides whether you can run this business profitably is "what can a professional manager earn here after I own it?" — and that is normalized EBITDA.

The gap between the two is the work the owner was actually doing. When a seller paid themselves $200,000, some of that was an owner's profit and some of it was the salary of the producer or account manager they were also being. You cannot assume that work happens for free once they leave.

Run that through a representative deal and the stakes are obvious. A seller presents $150,000 of net income plus a $200,000 salary add-back and calls it $350,000 of EBITDA; at a 3.0× multiple that is a $1,050,000 price. But the owner was the lead producer, and a replacement manager costs $100,000. The honest add-back is $100,000, the verified EBITDA is $250,000, and the true value at the same multiple is $750,000. The seller wasn't lying — they ignored the replacement cost. That gap is $300,000 of your money.

§ 02 · The add-back battlegroundValid, red-flag, and the proof each demands.

Most of the negotiation happens one add-back at a time. The seller proposes; you challenge; you settle. Repeat ten times and you have either uncovered real savings or been walked into overpaying. The discipline is sorting each claim honestly.

Add-backs that hold.

Expenses that genuinely will not recur under your ownership are fair game: a personal vehicle lease never used for client visits, a one-time legal settlement or a project-specific consulting fee, the owner's personal life-insurance premium paid by the business. Below-market rent counts too — but only with a third-party appraisal proving the gap and a commitment to stay in the building.

Add-backs to scrutinize.

"We can stop marketing and save $50,000" is not an add-back — marketing is how the book renews and grows; cut it and revenue follows. "My spouse does the bookkeeping for $60,000 and we can do it ourselves" is a cost shift, not a saving, unless you can name who does the work for free. Travel and meals need itemization to split client entertainment from personal. And contingency income should be normalized to a three-year average, never the peak year — a soft loss ratio makes those bonuses vanish.

Three levels of scrutiny.

Match your evidence demand to the claim. Low scrutiny for hard costs with invoices — a settlement agreement, equipment for a discontinued service. Medium scrutiny for partially personal expenses — allow the business share of a vehicle, prorate a home office by square footage. High scrutiny for soft estimates and projections — vendor "consolidation savings," staff-reduction claims, and any growth the buyer's own operating model would create. Soft estimates should rarely survive; the upside you bring is not the seller's to sell.

§ 03 · Two ratios that catch an engineered numberStress-test the whole package.

After you have a candidate EBITDA, two quick ratios tell you whether it is real or assembled.

A 45% margin can look like gold until you notice the owner hasn't hired a producer in three years and carriers are complaining about response times — the margin is borrowed from service quality you will have to restore, and the book will churn while you do. A comp ratio at 40% says the same thing from the other direction: someone is underpaid, and underpaid teams leave during a transition. When either ratio falls outside its band, stop accepting add-backs and make the seller explain the discrepancy.

§ 04 · The four-step rebuildFrom seller fiction to verified EBITDA.

The process that turns a summary into a number you can defend is short and repeatable. First, categorize every add-back by scrutiny level — hard cost, mixed-use, or soft estimate. Second, request documentation for each: invoice, receipt, statement, appraisal, or vendor quote. Third, apply the sustainability test — ask plainly, "will this cost recur after close?" — and if the honest answer is yes, it is not an add-back. Fourth, substitute a market-rate salary for every role the owner performed.

The output is your verified EBITDA, and three habits keep it honest. Get every add-back in writing — a verbal "we can cut that next year" you can't enter as a seller warranty is risk you own. Anchor the conversation with one question: when the seller says "we can eliminate that position," ask "at what salary, and who covers the work?" And carry the verified number straight into a forward pro-forma with real replacement salaries, conservative year-one attrition, and your debt service. If the agency can't cover the loan with a 1.25× cushion on those assumptions, the price is too high — regardless of how the add-backs were dressed.

◆