The diligence team does not start with your income statement. They start with your bank statements.
The first substantive procedure in most quality of earnings engagements is a proof of cash, which reconciles what your books say happened against what your bank says happened, month by month, across two or three years. Total deposits against recorded revenue. Total disbursements against recorded expenses. Not the ending balance, which any competent bookkeeper can make tie. The gross flows, both directions, which are much harder to fake and much easier to accidentally get wrong.
That choice of opening move tells you what the entire exercise is. A QoE is not a search for dishonesty. It is a test of whether a stranger, working only from documents you hand over, can rebuild your numbers and arrive where you said they would.
Most sellers who struggle through this process are not hiding anything. Their numbers are simply not reproducible by someone who does not already know the business. That is a different problem than fraud, it is far more common, and unlike fraud, it is fixable in advance.
What Is Actually Being Tested
Owners walk into diligence prepared to defend their business. The business is rarely the issue. What gets tested is the evidentiary chain behind every figure you have asserted.
Think of it as a stranger standard. Someone with no context, no relationship with your bookkeeper, and no willingness to take your word for anything receives a folder of documents and attempts to reconstruct your last three years. Every number that survives that reconstruction becomes real. Every number that cannot be traced back to a bank statement, an invoice, a contract, or a payroll register becomes a question, and questions in diligence do not resolve in the seller’s favor.
This is why “my accountant knows how that works” fails as an answer. It may be entirely true. It is also unverifiable, and unverifiable is functionally identical to wrong in this setting. The analyst cannot underwrite institutional knowledge.
Once you internalize the stranger standard, most of what follows becomes obvious. Preparation is not about polishing the story. It is about assembling the paper trail that lets someone else reach your conclusions without your help.
How the Review Actually Runs
Knowing the shape of the process removes about half the stress, because most of what feels alarming in the moment is simply the process working normally.
For SMB transactions, expect three to six weeks from kickoff to final report. It begins with an information request list, usually a spreadsheet running one hundred to three hundred line items covering financial statements, tax returns, bank statements, AR and AP detail, payroll registers, customer contracts, and organizational documents. Most of it is standard, and the volume is not a signal that anyone suspects anything.
From there: data room population, a period of analysis punctuated by follow-up requests, one or more management meetings, a draft report, a window for management response, and a final report delivered to the buyer.
Who is actually asking the questions
The staffing model matters more than sellers realize. A partner or director scopes the engagement and forms the ultimate view. A manager runs the workstream. One or two associates do the actual tie-outs, and they are the people generating the questions landing in your inbox.
Those associates are often two or three years out of school. They are competent, they are working from a standard program, and critically, they have no context for your business. When an associate asks why April’s revenue jumped 40%, they are not insinuating anything. They ran a monthly trend, saw an anomaly, and are required to explain it in the report.
The practical consequence: answer the question that was asked, provide the supporting document, and do not treat the inquiry as an accusation. Sellers who get defensive with associates burn days on friction that produces nothing, and the friction itself becomes an observation the manager passes upward.
The Procedures That Decide the Outcome
Four analytical moves carry most of the weight. Understanding what each one is designed to catch tells you exactly where to prepare.
Proof of cash, and why you should run it yourself first
This is the procedure that most often produces the finding that ends a deal, and it is the one sellers almost never anticipate.
A month-end bank reconciliation answers a narrow question: does the ending balance agree with the books on that date. It can tie perfectly while concealing an unrecorded deposit and an unrecorded payment that happen to net to zero. A proof of cash is wider. It reconciles the totals of money in and money out across the entire period, separately, so unrecorded activity has nowhere to hide.
What it commonly surfaces has nothing to do with wrongdoing. Sales tax collected and booked as revenue rather than a liability, which inflates the top line. Bank fees netted out of deposits so the gross receipt was never recorded. Interbank transfers between accounts that got coded as revenue or expense. Duplicate vendor payments. Merchant processor settlements recorded net of fees. Every one of these is an ordinary bookkeeping shortcut, and every one produces a gap between banked cash and reported revenue.
Kroll’s transaction advisory guidance notes that proof of cash reconciliations routinely surface sales tax booked as revenue, bank fees netted out of deposits, unrecorded interbank transfers, and duplicate vendor payments. None of these are fraud. All of them distort reported earnings.
Source:
Kroll, “The Critical Role of Proof of Cash Reconciliation in Financial Due Diligence”
Some divergence is normal and expected, since receivables, deferred revenue, and pass-through items all create legitimate timing differences. Practitioners generally treat differences in the range of three to five percent of revenue as unremarkable. Beyond that, the analyst starts pulling threads, and the burden shifts to you to explain a gap you may not have known existed.
The preparation is straightforward and almost nobody does it. Take twenty-four months. For each month, total every deposit across every bank account and compare it to recorded revenue for that month. Do the same for disbursements against recorded expenses. Where the numbers diverge, identify why and document the reason. If you cannot explain a variance, you have just found the thing the buyer’s team will find in week two, except you found it while you still have time to fix the underlying classification rather than defend it under pressure.
The EBITDA bridge and how adjustments get graded
Every proposed add-back gets sorted, informally, into one of three buckets.
Accepted adjustments are documented, clearly non-recurring or clearly discretionary, and supported by source material. Above-market owner compensation benchmarked against a comparable salary survey. A one-time legal settlement with the settlement agreement attached. A family member on payroll who does not work in the business, evidenced by the payroll register and an org chart.
Haircut adjustments are partially allowed. You claim $80,000 of owner discretionary expense, the analyst can substantiate $52,000, and the rest is disallowed for lack of support. This is the most common outcome for poorly documented add-backs, and it is quietly expensive.
Rejected adjustments are removed entirely. Pro forma adjustments draw the heaviest scrutiny here, and sellers frequently misunderstand the distinction. A normalizing adjustment removes something that happened but should not recur. A pro forma adjustment adds something that has not happened yet, such as the run-rate benefit of a contract signed last month or a cost you intend to eliminate post-close. Normalizing adjustments are routine. Pro forma adjustments are treated as forecasts, and forecasts do not get multiplied.
The operational lesson: an adjustment schedule that lists amounts is a starting position. An adjustment schedule where every line links to a specific document is a defended position. The difference between those two documents routinely runs into six figures of enterprise value.
The margin walk
Analysts decompose changes in gross margin across periods into their drivers: price, volume, mix, and cost. If margin improved 300 basis points last year, they want to know which of those four explains it.
The reason is simple. Margin improvement driven by a permanent price increase or a durable mix shift is sustainable and gets valued. Margin improvement driven by a one-time favorable input cost, a single unusually profitable job, or a customer that will not repeat is not. When a seller cannot explain their own margin movement with data, the analyst defaults to the conservative interpretation and treats the improvement as temporary.
Cut-off testing
They will examine the final two weeks of each fiscal period and the first two weeks of the next, looking for revenue recognized before it was earned or expenses pushed into the following period. In project-based businesses this extends to percentage-of-completion estimates, where a small change in estimated cost to complete moves recognized revenue meaningfully.
This is worth knowing not because most sellers do anything wrong here, but because the period around year end is exactly when a small business under pressure to show a strong year is most likely to have gotten aggressive without thinking of it that way.
The Two Things That Move More Money Than EBITDA
Sellers spend their preparation on add-backs. Meanwhile two other mechanisms are quietly determining what actually lands in the bank account, and both get almost no attention until it is too late to influence them.
The net working capital peg
Nearly every private-target deal includes a working capital adjustment, and it is the most common purchase price adjustment mechanism in the market.
Working capital purchase price adjustments now appear in more than 90% of private-target M&A transactions, up from roughly 50% a decade ago, according to SRS Acquiom’s Working Capital Purchase Price Adjustment Study. Among private-equity buyers, usage is effectively universal.
Source:
SRS Acquiom, M&A Working Capital Purchase Price Adjustment Study
Nearly every private-target deal includes a working capital adjustment. SRS Acquiom’s M&A Deal Terms research has put the figure above 96% of transactions, which makes it the most common purchase price adjustment in the market.
Here is the mechanic. The QoE establishes a normalized level of working capital the business requires to operate, called the peg or target. At closing, actual delivered working capital is measured against it. The purchase price moves dollar for dollar with the difference. Deliver $200,000 below the peg and your proceeds drop by $200,000.
The peg is not a fact. It is a calculation, and calculations have assumptions. Which months get included. Whether unusual months are excluded as outliers. How seasonality is handled. Whether the target is set on a trailing twelve month average or on a more recent run rate. A buyer generally benefits from a higher peg and a seller from a lower one, and the analysis proposing that number comes from the buyer’s advisor.
Sellers who prepare their own working capital analysis before signing an LOI walk into this with a position. Sellers who do not receive a number and are asked to agree to it. A trailing twelve month average can be countered with a forward-looking peg built on recent quarters and current customer terms, provided you have the monthly support file to justify it. Seasonal businesses in particular should be examining whether the averaging period disadvantages them.
Two things to know beyond the number itself. Collars are common, meaning a band inside which no adjustment occurs, which reduces exposure to small variances. And be alert to double counting: an item that reduced EBITDA in the QoE should not also reduce the purchase price through the working capital adjustment. That rule needs to be written into the agreement, because it does not enforce itself.
One warning. Buyers actively look for window dressing, meaning accelerated collections or stretched payables in the months before close that temporarily inflate delivered working capital. The detection method is a DSO and DPO trend analysis, and it is not subtle. Attempting it damages credibility at exactly the wrong moment.
Debt-like items
This is the mechanism that surprises sellers most, because the term itself sounds narrow and the application is broad.
In a cash-free, debt-free structure, funded debt gets paid off at close and reduces proceeds. Everyone expects that. What sellers do not expect is how many balance sheet items a buyer will characterize as debt-like and deduct on the same dollar-for-dollar basis: customer deposits, deferred revenue, accrued but unpaid income taxes, accrued bonuses, accrued paid time off, deferred compensation, capital and finance lease obligations, related-party balances, factoring arrangements, merchant cash advances, unfunded pension obligations, and litigation reserves.
Deferred revenue is the one worth watching closely. A subscription business collecting annually in advance is sitting on a balance the buyer will argue represents an obligation to deliver service without receiving new cash. Treated as debt-like, it comes straight off the purchase price. The counterargument exists, since prepayment also carries real economic benefit, but it has to be made, and it has to be made in the LOI rather than discovered in week four of diligence.
For construction and contracting businesses, this territory gets its own vocabulary. Retainage, underbillings, overbillings, mobilization advances, and work in progress all bear on both the working capital calculation and the debt-like analysis, and the treatment of each should be negotiated explicitly rather than assumed.
The total of these deductions frequently exceeds the value swing from disputed EBITDA adjustments. It receives a fraction of the preparation.
Where Books Quietly Betray Sellers
Four recurring problems, none of which involve anyone doing anything wrong.
Books that do not tie to tax returns is the most common by a wide margin. The pattern is familiar: the CPA makes year-end adjusting entries to prepare the return, and those entries never get pushed back into QuickBooks. Three years later the books and the returns tell measurably different stories, and the analyst comparing them has no way to know which is right. Reconciling book income to filed returns for three years, with the differences documented, is a weekend of work that prevents a genuinely damaging finding.
Multiple versions of the truth is the second. The confidential information memorandum says one revenue figure, the accounting system says another, and the tax return says a third. Each may be defensible in isolation. Together they signal that nobody has control of the numbers, and that impression is very difficult to reverse once formed.
Capitalized costs that should have been expensed inflate EBITDA and get caught routinely. Software development, internal labor, repairs treated as improvements. The analyst reviews the fixed asset additions schedule and tests whether the capitalization policy was applied consistently.
Related-party arrangements surface whether you disclose them or not. Below-market rent paid to an entity you own, a service agreement with a company owned by a family member, personal expenses in a vendor account. These are normal in owner-operated businesses and they do not kill deals. Discovering them without prior disclosure does.
How to Conduct Yourself
The analytical outcome matters. The behavioral read matters more than sellers expect, because the partner writing the report is forming a judgment about management credibility throughout, and that judgment colors how ambiguous findings get characterized.
Response speed is the clearest trust signal available to you. Requests answered within a day or two read as a company in control of its records. Requests that sit for a week read as a company that either cannot locate the documents or does not want to. The analyst cannot tell those two apart and will assume the less flattering one.
Never revise a number without explaining the revision. A changed figure with an explanation is a correction. A changed figure without one destroys confidence in every other figure you have provided, including the ones that were right.
Escalate methodology disputes rather than arguing them at the associate level. If you disagree that an adjustment should be disallowed, that is a judgment call the manager or partner owns. Litigating it with the associate who flagged it consumes days and changes nothing.
Disclose bad news before it is discovered. This is the single highest-leverage behavior in the entire process. A known problem raised by the seller in week one is a diligence item to be quantified. The identical problem found by the analyst in week four is a credibility event, and the buyer will reasonably wonder what else has not been mentioned. The finding costs you money. The surprise costs you the relationship, and sometimes the deal.
Use the management response window seriously. Most sellers treat the draft report as a verdict. It is not. There is a defined opportunity to respond, and rejected or haircut adjustments can be reinstated with documentation that was not available when the analyst made the call. Coming to that meeting with organized support has recovered real value in a lot of transactions.
What to Do Before They Arrive
Run your own proof of cash across twenty-four months, both directions, and document every material variance. This is the highest-return preparation available.
Reconcile three years of book income to filed tax returns and write down the explanation for each difference.
Rebuild the adjustment schedule so that every line item hyperlinks or attaches to its supporting document. Not a list. An evidence file.
Produce thirty-six months of monthly financials in a single consistent format, prepared under the same policies throughout. Format changes midstream generate questions that consume days.
Extract revenue by customer by month for thirty-six months into a flat file. They will ask for it. Having it immediately is a credibility event on its own.
Calculate monthly net working capital for twenty-four months, excluding cash and debt, and identify seasonality and outliers. This is your position on the peg, and it needs to exist before someone hands you theirs.
The Bottom Line
A quality of earnings review is a reproducibility test administered by strangers with no incentive to give you the benefit of the doubt. It is not an audit of your character, and treating it as one leads sellers to defend when they should be documenting.
The work that determines the outcome happens before the kickoff call. Proof of cash reconciled. Books tied to returns. Every adjustment supported by paper rather than explanation. A working capital position calculated and defended rather than received. Debt-like treatment negotiated in the letter of intent rather than absorbed in week four.
Businesses that do that work tend to find diligence uneventful, which is the objective. Businesses that do not spend three weeks answering questions about their own numbers from a position of disadvantage, and pay for the privilege in the final purchase price.
Where to Go From Here
The technical work of preparing for a QoE is learnable. The judgment calls are the hard part: which adjustments will actually hold, where the working capital position should be set, what needs disclosing early, and how to respond when the draft report lands with findings you disagree with.
At Frak Finance, this is the work: proof of cash and reconciliation, defensible adjustment schedules with the support attached, working capital analysis prepared from the seller’s side, and management through the review itself.
Schedule a free consultation and we’ll pressure-test your numbers before someone with a fee and a mandate does it for you.
