The Quality of
Earnings Report

What a QoE Is, and What It Is Not

A quality of earnings report is a focused accounting analysis of whether the profit a business reports is real, repeatable, and transferable. Owners frequently assume it is a light audit, and the distinction matters because the two answer different questions. Anders draws the line cleanly: a financial audit confirms GAAP compliance, while a quality of earnings report evaluates whether earnings are sustainable after a transaction closes. A QoE asks a commercial question that no accounting standard addresses. How much of this profit will still exist next year, under different ownership, without you in the building?

There is a second difference nobody mentions until it matters. An audit is an attest engagement performed under professional standards, and it ends in an opinion. A QoE ends in neither. Warren Averett states it directly: a quality of earnings analysis is not an audit and therefore no opinion is given. Midwest CPA goes further, noting that a QoE is a consulting engagement with no explicit AICPA standards on the processes or deliverables it must include, and that a person does not need to be a CPA to perform one.

Read that again, because it has a practical consequence sellers underrate. Two firms can both hand you something titled "Quality of Earnings" and have done substantially different amounts of work. The word on the cover is not a controlled term. When you compare quotes, you are not comparing price for a standardized product; you are comparing price for whatever that firm has decided the engagement includes. Ask for the procedures list, not the brochure.

The deliverable is a normalized EBITDA figure plus a schedule showing every adjustment made to get there from your reported earnings. That schedule is the document your price gets negotiated against. Owners tend to focus on the multiple because it is the number that sounds like it matters, but a half-turn of multiple on $2 million of EBITDA is $1 million, and so is a $500,000 haircut to EBITDA at a 2x difference in what survives. The adjustment schedule is where a large share of real negotiation happens, and it is the part most sellers have never seen before their own deal.

How large can the gap be? CliftonLarsonAllen writes that in owner-operated businesses, adjusted EBITDA can differ from reported EBITDA by 30% to 50% or more. That is CLA's characterization rather than a survey finding, and it is a QoE provider describing the value of QoE work. We quote it because it is directionally consistent with what we see, not because it is independently established.

Our guide to EBITDA and business valuation basics covers how adjusted earnings are constructed. This article covers what happens when someone independent tests that construction.

What It Costs, and What Scope You Are Actually Buying

Published cost figures for a sell-side QoE on a lower middle market business cluster around $25,000 to $50,000, with a typical timeline of four to six weeks from engagement to draft. Smaller engagements below roughly $3 million of EBITDA are quoted nearer $15,000 to $25,000, and larger or multi-entity businesses run from $50,000 to $75,000 and upward. The published spread is wider still at the edges. Morgan & Westfield puts the range at $5,000 at the low end for a minimal report on a small company to more than $100,000 for a complete report on a mid-sized firm, and gives a working timeline of 30 to 45 days depending on how fast management answers questions.

Those ranges deserve a caveat we would rather state than bury. Nearly every published QoE price you will find online is posted by a firm that sells QoE engagements. They are rate cards, not survey data, and the spread across sources is wide enough that some of the variation is positioning rather than genuine market difference. We use them because they are the best available, not because they are independent.

Full Scope, Limited Scope, and the Databook

Price varies partly because "QoE" covers at least three different products. Sellers who only ask what a QoE costs get quoted whichever one the firm prefers to sell.

A full scope engagement is the complete workstream: adjusted EBITDA, proof of cash, net working capital, revenue and margin analysis, balance sheet review, and concentration testing. A limited scope engagement narrows deliberately. Bridgepoint Consulting describes it as a streamlined version focused on specific financial areas rather than a full-scale review, used when the transaction is smaller or the parties want to expedite the close. DueDilio prices lite scope at $5,000 to $15,000 and full scope at $15,000 to $30,000 or more for very complex businesses, with lite scope completing in a few weeks against several weeks to months for full scope. Both are provider-side figures.

The third product is the databook: the underlying schedules and exhibits rather than a narrative report. A databook is cheaper and faster because nobody writes the analysis up. It is fine as an internal tool. It is weak as a document you hand a buyer, because the interpretation a narrative report supplies is exactly the part that defends your adjustments when someone attacks them.

Here is the trap. A limited scope report that omits net working capital leaves the peg entirely to the buyer's provider, and the peg moves your proceeds dollar for dollar. Saving $12,000 on scope to concede a $200,000 working capital argument is a bad trade, and it is a common one.

The commissioning and timing distinctions above follow the standard description used across accounting firms, including RKL, which makes the timing point bluntly: the advantage shifts from the seller to the buyer the moment a letter of intent is signed. The last two rows are our characterization, not a cited finding.

The cost sits inside a longer clock than most owners expect. How much longer is harder to state than it should be. A figure circulates widely in advisory commentary that due diligence in the $5 million to $10 million segment now averages 5.5 months, a record high, and it is routinely credited to IBBA and M&A Source Market Pulse. Calder Group is one firm reporting it that way. We went looking for it at the source and could not find it: neither the Q2 2025 nor the Q4 2025 Market Pulse release contains any diligence duration figure at all. So treat 5.5 months as an industry working assumption rather than a published statistic. What is not in doubt is the direction. Diligence takes months, not weeks, and every week of it is a week you are running your business while answering questions about it. Performance during diligence is itself something buyers watch.

Who This Article Is For

This is for owners of businesses roughly in the $2 million to $75 million revenue range, generally with at least $500,000 of adjusted EBITDA, who are either preparing to sell or already in a process and have just been told a QoE is coming. If you have never sold a business before, you are in the majority. IBBA and M&A Source reported that individual buyers accounted for 44% of lower middle market acquisitions in the fourth quarter of 2025, split 26% first-time buyers and 18% serial entrepreneurs, with private equity making about one fifth of acquisitions in that market. Most of the people on the other side of your deal have done this more often than you have, and the QoE is where that experience gap shows up most expensively.

What a QoE Actually Finds

Generic descriptions of QoE work are useless to a seller. What follows is the specific mechanism of each adjustment category, because knowing that recurring costs get scrutinized is worthless next to knowing exactly why yours will not survive. The category framework here follows the adjustment taxonomy published by Bonadio; the failure modes are our own.

Owner Compensation: The Most Expensive Misunderstanding

This is the error we see most, and it is definitional rather than dishonest. Two conventions exist for handling what the owner takes out of the business, and they produce very different numbers.

Under seller's discretionary earnings, the owner's entire compensation is added back to profit, on the logic that a buyer stepping into the owner's role captures the whole salary. Under EBITDA, only the portion above a market-rate salary for the work actually performed is added back, because a buyer still has to pay somebody to do the job.

Concretely: you take $300,000, and a general manager doing your actual job would cost $200,000. An SDE presentation adds back $300,000. An EBITDA presentation adds back $100,000. That is a $200,000 difference in earnings, and at a 5x multiple it is a million dollars of enterprise value that was never there. Owners who build a price expectation from an SDE-style add-back and then negotiate against an EBITDA multiple are wrong before the first meeting, and the QoE is where they find out.

Recurring Costs Dressed as One-Time

This is the most consistently rejected category. The test a QoE applies is not what the invoice is labeled; it is whether the cost appears again. A consulting engagement that shows up in three consecutive years is recurring, whatever the description says. Legal fees are non-recurring only if the underlying matter is genuinely resolved. A software implementation is one-time; the annual subscription that follows it is not.

The compounding damage matters more than any single rejected line. When a provider disallows two or three add-backs as improperly characterized, they begin testing everything else more aggressively, and legitimate adjustments start getting challenged because your schedule has lost credibility. An aggressive add-back schedule does not merely fail on its weakest items; it weakens the strong ones. Kreischer Miller frames the starting position from the buyer's side: without independent analysis, buyers are often skeptical of owner-provided add-backs, and when a buyer finds one error they begin wondering what else is hidden.

Pro Forma and Run-Rate: The Two Adjustments Buyers Trust Least

Most add-backs look backward. Two do not, and they are treated very differently from the rest.

A pro forma adjustment restates history as though a change that has already happened had applied for the whole period. Bonadio gives the standard examples: a rent increase or a staffing change that took effect partway through the year, projected across the full period so the historical numbers reflect the business the buyer is actually acquiring. BDO treats pro forma adjustments as a distinct category in working capital analysis for the same reason: applying a consistent methodology retroactively across historical periods.

A run-rate adjustment annualizes recent performance. Three good months become twelve. The logic is not crazy, and occasionally it is right. It is also the single easiest adjustment to abuse, and everyone downstream of you knows it.

The ranking that matters is this. An adjustment tied to a signed contract, an executed lease, or a termination that already occurred is close to unarguable, because the document exists and the cost change is committed. An adjustment tied to a decision made but not yet fully reflected in results is negotiable. An adjustment tied to a trend you expect to continue is a forecast, and buyers do not pay historical multiples on forecasts. Run-rate loses hardest because it is the only one of the three where the evidence is the thing being claimed.

If you want a run-rate adjustment to survive, give it a document rather than a trajectory. A signed customer contract with pricing and term beats six months of rising revenue in the same account, every time.

Revenue Recognition and Cut-Off

Cut-off testing checks whether revenue landed in the period it belongs to. Recording a December shipment that delivered in January inflates the year buyers are underwriting, and a QoE will move it. This is rarely deliberate and frequently expensive, because owners on cash-basis or hybrid books often have no consistent policy at all. Where revenue is recognized over time, or where deposits and prepayments are involved, the analysis gets harder and the range of defensible answers gets wider — which is exactly the situation where having your own report first is worth the money.

Quality of Revenue, Which Is Not the Same Thing

Quality of earnings asks whether the profit is real. Quality of revenue asks whether the revenue that produced it will still be there in three years. These are separate workstreams, and sellers routinely prepare for the first and get ambushed by the second.

RSM lists revenue, margin and customer analytics alongside quality of earnings as distinct components of financial due diligence, not as a subset of it. In practice that means someone is going to rebuild your revenue by customer, by year, and look at what happened to each cohort.

The specific tests are unglamorous and hard to argue with once run:

• Cohort retention. Group customers by the year they were acquired, then track what each group spent in every subsequent year. Growth that comes from new logos replacing churned ones looks identical to durable growth on a revenue line and nothing like it in a cohort table.
• Logo churn versus revenue churn. Losing 15% of customers who represent 3% of revenue is a different business from losing 3% of customers who represent 15% of revenue. The top-line growth rate hides both.
• Concentration. Not just the largest customer, but the top five and top ten as a share of revenue, and whether those relationships are contracted or habitual.
• Recurring versus reoccurring. A contract that auto-renews is recurring. A customer who happens to order every year is reoccurring, and it is worth materially less because nothing obliges them to come back.

That last distinction costs sellers more than any other item on this list. Owners describe reoccurring revenue as recurring in good faith, because from inside the business the cash arrives on the same rhythm either way. A buyer paying a multiple for contracted revenue and discovering habit will reprice, and will be right to.

Concentration is where quality of revenue stops being an accounting exercise and starts moving price. Our guide on what buyers look for in an acquisition target covers the thresholds in detail.

Net Working Capital, and How the QoE Sets Your Peg

This is the workstream sellers understand least and pay for most.

Net working capital analysis usually rides inside the same engagement as the QoE, and BDO identifies it as one of the key areas of financial due diligence alongside quality of earnings and a debt and debt-like items analysis. The output is the peg: the baseline amount of working capital you are expected to leave in the business at closing. BDO describes the peg as typically an average of normalized or adjusted net working capital for the latest trailing twelve months, though the averaging period may be shortened to six or even three months where that better represents the current state of the business.

Then the arithmetic gets unforgiving. If closing working capital exceeds the peg, BDO notes the buyer may pay the seller the excess dollar for dollar. Fall short and the purchase price drops the same way. There is no multiple applied and no negotiating room once the mechanism is in the purchase agreement. A peg set $250,000 too high costs you $250,000, quietly, at closing.

Two things drive the peg your way or against you. The first is the averaging window, because a business with seasonality will produce a very different twelve-month average than a three-month one, and whichever period the buyer's provider selects will not be an accident. The second is classification. BDO notes that certain current liabilities are better treated as debt-like or long-term, listing accounts payable aged over one year, bonuses and accrued profit sharing, and change-in-control payments. Every item reclassified from working capital to debt reduces your proceeds, because debt comes off the purchase price at closing.

The peg deserves more attention than it usually gets, and we have given it a full treatment in our guide to the working capital peg in M&A. If you read one thing after this article, read that one.

Reserves, Inventory, and Related-Party Arrangements

Reserves get tested against history. An inventory reserve that has not moved in four years while inventory aged is an understated cost, and a warranty reserve that does not reflect actual claims is the same problem in a different account. Both raise reported EBITDA today by deferring a real cost.

Related-party arrangements are where the largest single adjustment usually appears, and rent is the most common. If your operating company pays rent to an entity you own, that rent is almost certainly not at market. Below-market rent inflates the operating company's earnings; above-market rent understates them. A QoE normalizes it to market either way, because after closing the buyer pays a real landlord or pays you under a real lease. Sector guides on our site cover how large this can get: it is the dominant adjustment in childcare and daycare valuations, where owners frequently own their buildings.

Proof of Cash, and the Limit Nobody Advertises

A proof of cash reconciles what the books say against what the bank shows, unwinding movements in receivables and payables to test whether money actually moved as recorded. It is genuinely good at catching timing problems, transfers between accounts that obscure activity, and nonstandard journal entries.

It is worth being precise about what it does not do, because the marketing around QoEs frequently implies otherwise. FORVIS Mazars states directly that proof of cash alone cannot tell you whether fraud exists. What it does is surface unusual transactions that require further investigation. A fraud constructed so the accounting stays internally consistent — fictitious revenue matched by fictitious receivables — reconciles cleanly. A QoE is a test of earnings quality, not a fraud examination, and any provider implying otherwise is overselling.

Carve-Outs: When You Are Selling a Division, Not a Company

Selling one division out of a larger business changes the QoE from a test of reported numbers into a construction of numbers that never existed. The division has no standalone financial statements, because it was never a standalone anything.

EisnerAmper describes the core problem: buyers must estimate stand-alone costs, meaning the costs needed to run the business without the parent entity, which involves significant assumptions that could differ from actual post-transaction costs. Accounting standards require only a reasonable allocation method, such as revenue or activity levels, and EisnerAmper notes those allocations might not reflect true future costs, potentially making the business seem more profitable than it actually is.

The direction of the error is consistent, and it runs against the seller. EisnerAmper puts it plainly: expect historical cost allocations to result in fewer full-time employees and insufficient IT, marketing, HR, and legal support for the business unit to function independently. The carve-out looked profitable partly because the parent was absorbing costs nobody charged it for.

So the stand-alone cost adjustment is usually negative, often materially, and it is the buyer who quantifies it. The seller's defense is to do that arithmetic first: what does this division actually pay for its own accounting, insurance, IT, and HR on day one, and which of those services will the parent supply temporarily under a transition services agreement rather than permanently? A transition services agreement that expires in twelve months is not a cost saving. It is a deferral, and a buyer's provider will model the cliff at the end of it.

Which Add-Backs Actually Survive

Below is how the common categories tend to hold up. The categories follow the adjustment taxonomy published by Bonadio and the diligence scope described by RSM and Warren Averett. The survivability column is our characterization from doing this work, not a published statistic.

One pattern runs through the whole table. Adjustments backed by a document survive; adjustments backed by an explanation do not. That is the entire rule, and it is why the preparation work described at the end of this article is worth more than any argument you make during the engagement.

What Happens When the Lender Reads Your QoE

Most sellers picture one audience for the QoE. There are two, and the second one is stricter.

If your buyer is borrowing to fund the purchase, and in the lower middle market they usually are, a credit provider reads the same report and reaches a more conservative number. The lender is not pricing an asset they hope will grow. They are sizing debt against cash flow they need to be there in a downturn, and their definition of EBITDA is a contract term rather than an analytical judgment.

The mechanics are visible in how credit agreements are drafted. Sidley Austin notes that lenders often impose caps, frequently shared caps expressed as a percentage of EBITDA, on projection-based or non-recurring adjustments, along with time-based restrictions, specifically in an attempt to reflect actual operating cash flows of the business. Sidley also reports that a typical leverage covenant in a direct lending transaction may be set with a 25% to 35% cushion to the EBITDA projected in a sponsor or borrower model.

Osborne Clarke, writing on loan agreement practice, describes the specific controls lenders apply to forward-looking add-backs: a hard cap set against a percentage of EBITDA, senior management certification of anticipated synergies that ratchets upward as the adjustment grows, independent third-party certification before the final cap, and a common requirement that synergies be realized within 12 months of completion. Lenders also typically accept cost synergies while rejecting revenue synergies outright. Frost Brown Todd describes the same negotiation from the drafting side, noting that the goal of a credit facility EBITDA definition is a market-accepted approximation of the borrower's consistent operating cash flows.

Two consequences for a seller. First, an adjusted EBITDA that survives the buyer's QoE can still be trimmed by the lender, and if the trim reduces how much debt the buyer can raise, the buyer either finds more equity or comes back to you on price. Second, aggressive run-rate and synergy add-backs are worth even less than the QoE section above suggests, because they hit a second filter with contractual caps rather than professional judgment behind it.

If the buyer is using SBA financing, the underwriting is more prescriptive still. That is worth knowing before you assume an individual buyer's offer is as flexible as a private equity buyer's. Our comparison of a strategic buyer versus a private equity buyer covers how buyer type changes what the financing will tolerate.

Reliance: Whose Report Is It, Legally?

A question almost no seller asks, and it determines whether your $40,000 report does any work for the other side.

A QoE is delivered to the party that engaged it. Anyone else reading it is reading a document they have no legal right to depend on, unless the provider issues a reliance letter naming them. Without one, the buyer, the buyer's lender, and any representation and warranty insurer are looking at your report as information rather than as something they could act on and have recourse over.

The convention in auction processes is well established. A&O Shearman, writing on vendor due diligence practice, notes that draft reports are made available to prospective bidders on a non-reliance basis, with reliance extended only to the successful bidder once a definitive purchase agreement is signed, and typically subject to a cap on the provider's liability that falls away only in cases of gross negligence, recklessness, or fraud. That analysis describes large-cap auction practice rather than lower middle market deals, but the structure carries down.

Three practical points. Reliance is priced and negotiated at engagement, not afterward, so ask about it when you hire the firm rather than when the buyer requests it. Reliance comes capped, and our experience is that the cap is small relative to what a bad number can cost you, though the level is negotiated deal by deal and we know of no published benchmark for it. And extending reliance does not eliminate the buyer's own work: Morgan & Westfield observes that most buyers will retain a firm to perform a buy-side QoE analysis even when the seller has had one.

Which reframes what a sell-side QoE is for. It is not a substitute for the buyer's diligence and it will not be treated as one. Its value is that it tells you what the buyer's provider is going to find, early enough that you can fix it, explain it, or price it in before anyone has exclusivity.

The Claims About QoEs That Do Not Survive Checking

We spent a meaningful part of the research for this article trying to verify the statistics that appear in nearly every QoE article online. Most of them could not be traced to a primary source. Since those numbers are frequently quoted at sellers to justify a fee, the failures are worth naming individually. Where a later check has since closed a gap, we have said so in the table rather than leaving the original verdict standing.

We are not arguing that a sell-side QoE is a bad idea. We are arguing that the case for one should be made on a mechanism you can verify rather than a statistic you cannot. The defensible case is this: a QoE finds what it finds regardless of who commissions it. Commissioning it yourself changes when you learn the answer, and therefore how much leverage you have when you respond. Finding a $400,000 revenue recognition problem while five buyers are competing is a fixable inconvenience. Finding it after one buyer holds a signed letter of intent and exclusivity is a price renegotiation you will mostly lose. That argument does not need a percentage attached.

The GF Data split above is the strongest evidence for that framing, and it arrived from a direction we did not expect. If the valuation premium concentrates in deals above $50 million of enterprise value and largely disappears below it, then for most owners reading this the multiple argument is simply not the reason to do it. What does not depend on deal size is the sequencing. A problem found before you go to market is a problem you can fix, document, or price in. The same problem found under exclusivity is a discount. That holds at $8 million of enterprise value exactly as it holds at $80 million, which is more than can be said for the 0.4 of a turn.

The Axial figure above is the closest thing to evidence we located, and it is worth being careful about what it does and does not show. That roughly a fifth of broken 2025 LOIs in one platform's deal flow traced to QoE-identified EBITDA discrepancies tells you the finding is consequential. It does not tell you a sell-side QoE would have prevented any of them, because nobody tracked the counterfactual. The honest inference is narrow: this is where deals break, so this is where preparation pays. Not: buy this product and your deal closes.

How This Plays Out by Sector

• Childcare and early education. Rent normalization dominates, because owners so often own the buildings. See childcare and daycare valuation multiples.
• Managed IT services. The scrutiny lands on revenue recognition across contract terms and whether recurring revenue is genuinely contracted and assignable. See MSP valuation multiples.
• Industrials and manufacturing. Inventory reserves and the line between maintenance and growth capital expenditure carry the analysis, since EBITDA ignores reinvestment a manufacturer cannot avoid. See our guide to manufacturing M&A advisors.
• Business services. Contracted versus project revenue, and whether customer relationships survive the owner's departure. See our guide to business services M&A advisors.
• Across all of them, customer concentration surfaces in the QoE even though it is not an accounting adjustment, because it determines how much of the earnings a buyer will treat as durable. Our guide on what buyers look for in an acquisition target covers the thresholds.
• Also across all of them, the working capital peg is set inside the same engagement and moves your proceeds dollar for dollar. Seasonal businesses should read our guide to the working capital peg before agreeing to an averaging period.

Preparing So the QoE Does Not Cost You Money

The work that protects you happens well before anyone is engaged. Move to accrual-basis financials and close your months on a consistent schedule. Stop running personal expenses through the business at least a full year before going to market, so the year buyers underwrite is already clean rather than heavily adjusted. Document each add-back to an invoice when it occurs, rather than reconstructing the story a year later under deadline. Separate related-party arrangements and find out what market rent on your building actually is before a buyer tells you.

Three items sellers skip that cost real money. Build a customer-level revenue file going back three years, because someone will build it anyway and you want to see the cohorts first. Understand your own working capital cycle across a full year, so you can argue for an averaging period rather than accept one. And get a market rent opinion on any building you own and lease to the company, since that single adjustment is frequently the largest in the report.

None of that is fast, which is why it belongs in the window described in our guide on when to start exit planning, and why it is worth understanding how a lower middle market sale process unfolds before you are inside one. If you are still choosing who represents you, our guide on how to choose an M&A advisor covers what to ask about diligence preparation specifically.