What Add-Backs Survive Diligence

Key Takeaways

The test for any add-back: would this expense actually disappear under new ownership?
Excess owner comp is usually the biggest adjustment, but only if you size the real replacement cost
Personal expenses need logs and clean separation. No documentation means no add-back
"One-time" only counts if it is truly isolated. Patterns kill the add-back
A small EBITDA haircut becomes a big price haircut at a multiple

Most add-back schedules are optimistic. That does not automatically mean the seller is lying. It means they are looking at the P&L through the lens of how they ran the business, not the lens of what a lender will finance.

Your job is simple: turn the add-back schedule into a number you would defend to (i) yourself, (ii) a QoE team, and (iii) a bank. If you cannot defend it, it is not real earnings.

The only question that matters

Would it actually go away after close?

For every line item, ask:

Would this expense exist if the owner sold the company and walked away?
If not, what evidence proves it disappears or normalizes?

If the answer requires a speech, expect the add-back to get rejected.

Owner comp is where the dollars are

Size the adjustment without underestimating replacement cost

Owner compensation is usually the largest add-back. It is also the easiest place to overpay because buyers underestimate what it takes to replace the owner's workload.

Example: the seller takes $400K total comp. They claim a CEO is worth $150K, so they add back $250K.

That might be true. Or it might be a story.

If the owner is running operations, selling, and handling finance, you are not replacing "a CEO." You are replacing multiple roles.

A practical way to diligence it:

List what the owner does week to week (sales calls, quoting, scheduling, payroll, collections, vendor management).
Price those roles at market, in your geography.
Decide what you need on Day 1, not what you hope to do by Day 365.

If your real replacement plan costs $280K, the add-back is $120K, not $250K. At a 4x multiple, that difference is $520K of price.

Personal expenses die without paper

If you cannot prove it is personal, it stays in earnings

Some items are clean. A box truck that moves equipment is a business expense. No add-back. A personal SUV, personal travel, or a family member "advisor" salary can be legitimate add-backs, but only with proof.

Related-party rent is another common "adjustment." Treat it as a normalization, not magic. If rent is below market, you should assume it goes up post-close. If rent is above market, you still need comps and a path to a market lease.

Where these add-backs go to die:

Vehicles with no usage logs
Phone and meal reimbursements with no allocation
Club dues labeled "client entertainment" with no meeting notes or client names

Simple rule: no documentation means it stays in EBITDA.

If you want to move fast, ask for the support up front: general ledger detail, credit card statements for the owner period, vehicle logs, payroll detail for family members, and the lease if real estate is related-party.

One-time only counts once

Non-recurring is a claim. You still have to validate it.

A lawsuit settlement from three years ago might be a clean add-back. A string of settlements is a pattern and should stay in the run rate.

Same with write-offs. A big inventory write-off looks "one-time" until you learn it is happening every year, just in different forms. At that point it is a normal expense.

What you want to see:

A clear root cause
Evidence it was fixed (not just paid for)
Documentation that ties the cost to a finite event

A note on COVID: 2020 to 2021 noise should not still be driving trailing twelve-month add-backs in 2026. If "pandemic impact" is still in the schedule, treat it as a high burden-of-proof item.

Add-backs that usually fail

When you see these, assume the schedule is aggressive

These adjustments get rejected in most deals:

"We could cut costs but choose not to." That is upside, not earnings.
Deferred maintenance. The spending does not disappear. It usually gets worse post-close.
Customer concessions. That is pricing power being revealed, not a temporary discount.
Management cuts. If you need to cut to hit the number, you are paying for a turnaround you have not executed yet.

If the schedule leans on these, discount everything else until proven.

What a QoE haircut looks like

Small changes, big dollars

A company is marketed at $800K adjusted EBITDA. QoE rejects:

$150K of owner comp add-back because the owner is full-time and replacement cost is higher than claimed
$40K of vehicle expenses due to no logs
$75K of "one-time" legal costs because similar issues show up repeatedly

Adjusted EBITDA becomes $625K. That is a $175K haircut. At a 4x multiple, that is a $700K valuation swing.

This is why you do your own haircut case before LOI. You want the first surprise to happen in your spreadsheet, not in the bank's credit memo.

What's Next

Before you sign an LOI:

List every add-back and label it: accept, reject, or verify.
Build a haircut case where the weakest 20% to 30% get removed.
Size your offer and leverage to the haircut case, not the seller's ceiling case.

What this does not cover: normalized capex and working capital. Those can change your real cash flow even if adjusted EBITDA is correct.

Use the Earnings Calculator

Itemize add-backs, run a haircut case, and see how your valuation moves when the weakest adjustments get removed. Use it before LOI so you know the number you are actually buying. Output: adjusted EBITDA by scenario.

→ Earnings Calculator

Sources

U.S. Small Business Administration, SOP 50 10 (Version 8, effective June 1, 2025)
CBIZ, "Sell-Side Quality of Earnings: A Critical Part of Due Diligence"
Moss Adams, "Quality of Earnings Report in Pre-Sale Due Diligence"
BDO, "Purchase Price Wars: EBITDA vs. Adjusted EBITDA and Why It Matters"