Add-backs and what doesn't survive Quality of Earnings.
Why the seller's adjusted EBITDA almost always shrinks during diligence.
The single largest source of post-LOI renegotiation is the gap between what the seller's investment banker presents as adjusted EBITDA and what survives the buyer's quality-of-earnings review. This is a working seller's guide to that gap: what add-backs are, the taxonomy of legitimate vs theatrical, how the QoE team tests each one, and how to walk in with normalizations that hold.
Not sure where your business would break in diligence? Start with the Self-Assessment.
What add-backs actually are.
When a business is sold, the price is almost always expressed as a multiple of earnings — and the earnings number that drives the price is not the EBITDA reported in the financial statements. It is an adjusted EBITDA, normalized for items the buyer wouldn't inherit. The adjustments are called add-backs. They are the practical reality of how middle-market and lower-middle-market companies actually trade.
The concept itself is legitimate. Owners run businesses with a particular financial structure that reflects ownership, not enterprise economics — they pay themselves more than a CEO would cost, they expense personal items, they own the building, they have family on payroll. None of this would be true under institutional ownership, and the buyer needs to underwrite the post-close economics, not the seller's historical lifestyle structure.
The problem is that the add-back conversation is almost always asymmetric. The seller's investment banker prepares an “adjusted EBITDA” with every normalization that makes the business look strongest. The buyer's quality-of-earnings team then reviews each adjustment under a stricter standard. The gap between what gets presented and what survives is commonly the largest single source of post-LOI renegotiation in middle-market transactions.
Sellers who understand the QoE team's standard before they go to market preserve the headline number. Sellers who learn it during diligence don't.
Eight categories. What survives QoE — and what doesn't.
Almost every add-back in a middle-market transaction falls into one of these eight categories. In each, there is a credible version and a theatrical version. The QoE team's job is to identify which is which.
Owner compensation normalization
What survives QoE. Adding back the gap between what the owner is paid and what a market-rate CEO would actually cost the business. Documented with a market-comp study or comparable benchmarks. Adjusted downward to reflect the realistic post-close compensation structure.
What doesn't. Adding back the entire owner salary as if the business will be run without one. Or adjusting to a market rate so low it can't realistically recruit a replacement. Both fail because the buyer has to fund a real CEO.
Family members on payroll
What survives QoE. Adding back the compensation of family members who genuinely do not work in the business — or whose role can be eliminated post-close without operational consequence — supported by a role analysis and a hiring plan.
What doesn't. Adding back the salary of family members who do real work, because the role gets backfilled at the same or higher cost. The buyer normalizes to what it actually takes to replace the labor.
Personal expenses run through the business
What survives QoE. Adding back genuinely personal expenses that were charged to the business — provided they are individually documented, isolated to specific transactions, and the pattern is bounded. A clearly personal vehicle. A specific country club. A definite personal trip.
What doesn't. Sweeping categories of expenses that mix business and personal use. Travel that doubled as family time. Meals that were partly entertainment. Vehicles that were partly business. Add-backs without granular documentation get challenged across the category, not item by item — and the pattern alone damages credibility on every other adjustment.
“One-time” expenses
What survives QoE. Genuine one-time costs: a specific litigation settlement with documented resolution, a one-time facility move, a discrete acquisition-integration cost, a specific systems implementation. Each one verifiable, bounded, and clearly non-recurring.
What doesn't. “One-time” expenses that recur every year. Legal fees that show up annually but are claimed as one-time when they happen. Severance treated as one-time when the business has a pattern of turnover. IT upgrades treated as one-time when they cycle every three years. The QoE team looks at three to five years of history; recurring patterns surface.
Related-party rent normalization
What survives QoE. Adjusting above-market rent paid to owner-held real estate down to market rent (credit back to EBITDA) — supported by a third-party market study and a clean lease structure the buyer can inherit.
What doesn't. Adjusting below-market rent up to market when the seller wants to keep the real estate, because the buyer will require a new lease at market terms anyway. Or adjusting above-market rent down without offering a clean lease at the adjusted rate at close.
Discontinued products, customers, or lines of business
What survives QoE. Removing genuinely discontinued operations from the trailing financials — supported by a clear discontinuation date, no residual revenue, and no associated continuing obligations. The buyer underwrites the business they're buying, not the business that no longer exists.
What doesn't. Removing a customer who left because of a quality dispute or a service problem and calling it “discontinued.” That isn't a normalization — that's a churn event the buyer needs to understand for what it actually was.
Pro-forma adjustments for acquisitions
What survives QoE. Trailing-twelve-month financials that include a tuck-in acquisition's full-year results when the acquisition closed mid-period — supported by clean pre-close financials from the acquired entity and integration cost transparency.
What doesn't. Pro-forma adjustments that assume synergies the seller hasn't actually captured. The buyer credits what's been earned, not what's been hoped for.
“Run-rate” adjustments to a recent quarter
What survives QoE. Almost nothing. Run-rate adjustments — annualizing a single strong quarter to claim a higher EBITDA — are the single most aggressively discounted form of add-back. Buyers underwrite trailing-twelve-month performance.
What doesn't. A strong Q4 annualized into a full year of implied EBITDA. A new customer that just signed treated as if it's already produced twelve months of revenue. New pricing implemented mid-year treated as if it covered the entire year. The QoE team rejects these categorically.
How the QoE team actually tests an add-back.
The buyer's QoE team is not adversarial in posture — it is methodical. Each add-back gets four tests, and the ones that survive are the ones that pass all four.
A “one-time” cost that appears two years running is recurring
Owner comp only adds back at a market replacement rate
Undocumented add-backs don’t survive contact with QoE
The aggregate has to look conservative, not engineered
The recurring-pattern test.
A “one-time” expense that shows up two years in a row is recurring. The QoE team looks at three to five years of history specifically to test recurrence. Expenses that consistently appear — even if their specific cause varies — get rejected as one-time. Legal fees that come from different cases each year are still recurring legal fees.
The role-replacement test.
For owner compensation, family-payroll, and any role-related adjustment: what does it actually cost to replace the role? The QoE team asks for the job description, asks who would do this work post-close, and benchmarks against market data. Add-backs that imply work disappears under new ownership get challenged.
The documentation test.
Vague add-backs get rejected. Specific, individually-documented add-backs survive. An add-back for “owner travel” without itineraries gets discounted; an add-back for three specific trips with documented personal purposes survives. Granularity is credibility.
The pattern test.
The QoE team is reading every add-back together, not in isolation. If add-backs in one category are sloppy or aggressive, every other add-back gets a tighter standard applied. The seller who presents twelve credible add-backs survives QoE largely intact; the seller who presents twelve credible plus three theatrical ones often loses ground on all fifteen because the pattern signals strategic posture rather than honest normalization.
Why theatrical add-backs cost more than they could ever add.
A theatrical add-back doesn't just fail on its own merits. It changes the buyer's posture across the entire deal. The buyer who arrives at the QoE conversation expecting honest normalization, and finds creative accounting instead, shifts from “how do we get this done” to “how protected do we need to be.”
The cost shows up in three places. Reps and warranties tighten — the buyer wants more protection because they've seen the seller stretch. Escrow goes up — the buyer wants more held back. Earn-outs lengthen — the buyer wants more of the consideration contingent on actual performance rather than presented numbers. None of these are recoveries of the rejected add-back itself; all of them are surcharges layered on top of the price reduction the rejected add-back already caused.
The math almost never works for the seller. An aggressive add-back that gets rejected costs the multiple-times-the-rejected-amount in price plus the cumulative cost of tighter deal terms. The honest version of the same number — leaving out the theatrical part — costs only the difference between the theatrical and the credible amount, and preserves negotiating posture across everything else.
The discipline is conservative add-back presentation. Surface the normalizations that will survive QoE with documentation already in hand. Leave the theatrical ones out. Let the buyer's diligence find more strength than the marketing book promised, not less.
How to walk in with add-backs that hold.
Run a QoE-style review on your own EBITDA before going to market. Take each potential add-back through the four tests. If it doesn't pass all four, leave it out. The presented adjusted EBITDA should be a number that survives external review, not a number designed to maximize the marketing book.
Document every add-back individually, with source records. Vague categories are discounted; specific transactions are credited. A spreadsheet of every adjustment with date, amount, source document, and reason is the standard. Anything less surfaces as a finding.
Reconcile add-back history across the trailing three-to-five years. The QoE team will. If your add-backs in year three don't reconcile to your add-backs in year one — different categories, different amounts, different documentation — that surfaces as a credibility issue across the whole package.
Normalize owner compensation to a realistic post-close number. Not zero. Not the absolute minimum. The actual cost of a CEO who can run this business under institutional ownership. Aggressive normalizations get rejected; realistic ones get credited.
Treat family payroll, related-party rent, and personal expenses with extra rigor. These are the three categories where QoE teams have seen the most theatrical adjustments and where they apply the most skeptical standard. Documentation, role analysis, and market-rate benchmarks aren't optional — they're the price of admission.
The seller who presents adjusted EBITDA that survives QoE essentially intact is the seller who keeps the headline price. The seller who watches their adjusted EBITDA shrink by 10–25% during diligence almost always watches the price shrink with it — and frequently sees the deal structure shift on top of that.
What sellers most often ask before they go to market.
- What is an add-back in a Quality of Earnings analysis?
- An add-back is an adjustment to reported EBITDA that reflects an expense or income item the seller argues is non-recurring, owner-discretionary, or otherwise not representative of go-forward earnings. The buyer's QoE team evaluates each proposed add-back against four standard filters: recurrence, owner-vs-business, documentation, and ongoing-need.
- What add-backs survive Quality of Earnings?
- Add-backs that are documented, non-recurring, and unrelated to ongoing operations tend to survive — confirmed one-time legal settlements, completed transaction costs, related-party compensation differentials with documented market comparables, and discontinued product lines with verifiable cost separation. Owner-discretionary categories like personal vehicle costs and family payroll commonly hold when documented. Aggressive normalizations of recurring costs typically don't.
- How does QoE differ from an audit?
- An audit verifies that financial statements comply with accounting standards. A Quality of Earnings analysis tests whether the reported earnings represent durable, transferable cash flow under new ownership. The two examine the same financials but answer different questions — and a clean audit does not guarantee a clean QoE outcome.
- Why does adjusted EBITDA often shrink during diligence?
- Buyer QoE teams apply tighter filters than sellers commonly anticipate. Add-backs the seller treated as standard get rejected on documentation, recurrence, or ongoing-need grounds. Revenue-recognition timing gets re-pulled. Run-rate assumptions get walked back. A modest revision to adjusted EBITDA at five-to-eight-times multiple is the single most common retrade trigger in middle-market deals.
- Can a seller commission their own Quality of Earnings report?
- Yes — and many sophisticated sellers do, in the months before going to market. A seller-side QoE applies the same filters the buyer's team will apply, surfaces the add-backs that won't survive, and lets the seller walk into the LOI conversation with a normalized EBITDA picture that diligence is unlikely to revise downward.
Self-Assessment tells you what might break. Gap Review proves whether the documents support the same story.
See where your adjusted EBITDA actually stands.
The Self-Assessment grades your business against the same ten-axis instrument a buyer's diligence team applies — and surfaces the add-back patterns that would shrink your presented EBITDA before the QoE team finds them. Read your real position on your timeline, not theirs.