From 6 weeks to under 60 minutes: financial diligence, automated.

An EBITDA add-back is an expense added back to reported earnings because it is non-recurring, non-operational, or will not continue after the transaction closes. Add-backs raise adjusted EBITDA and therefore raise valuation.

Owner compensation is the standard example. A founder paying themselves $500,000 in a business that would pay a professional manager $200,000 creates a $300,000 add-back, because that excess cost disappears under new ownership. The logic is forward-looking, which is the source of most disputes: an add-back is valid only if the buyer genuinely will not incur the cost. Add-backs are the most contested element of most lower middle market negotiations.

A legitimate add-back is traceable in the general ledger, documented, and genuinely will not recur under new ownership. All three conditions have to hold, and most disputed add-backs fail on the second.

The five tests a buyer should apply to every line:

• Traceable. The item ties to specific journal entries and a specific amount, not to a management estimate.
• Documented. An invoice, contract, settlement agreement, or payroll record supports it.
• Non-recurring in fact. The item appears in one period only, rather than reappearing in a different form every year.
• Not operationally required. The business can run without the cost after close.
• Consistently treated. The same category of expense is handled the same way across all periods presented.

An add-back that fails the documentation test is an assertion. An add-back that fails the non-recurrence test is usually a normal operating cost that has been relabeled.

AI can do most of the data work: ingesting general ledger and trial balance files, normalizing charts of accounts, reconciling records, detecting anomalies, and proposing EBITDA adjustments with the supporting entries attached. It cannot replace professional judgment or attestation.

What it does well is high-volume, rule-consistent analysis across a complete dataset. What it cannot do is interview management, negotiate an adjustment with a seller’s advisor, interpret contractual nuance, or take professional responsibility for a conclusion. The realistic model is sequencing rather than substitution: use software to establish the factual picture fast, then apply human judgment where judgment is what is actually needed.

Yes. General ledger analysis is the strongest application, because the task is high-volume pattern recognition over structured data.

Software reads a full ledger regardless of size, standardizes account naming, groups transactions by vendor and category, trends accounts monthly, and identifies entries that deviate from established pattern. It handles cash-basis books and multi-entity structures where consolidation is manual. On one Quebec mandate, AddBack processed 177 documents and more than 130,000 accounting entries, and arrived at a normalized EBITDA within $2,000 of the accounting firm’s databook.

No, and buyers should be skeptical of anyone claiming otherwise.

A QoE report carries professional responsibility. Lenders, private credit funds, and investment committees rely on it because a named firm stands behind the conclusions. Software carries no such accountability, cannot conduct management interviews, and cannot exercise judgment on the negotiated questions that determine final adjustments.

The useful role is upstream. Establish the financial picture before a QoE engagement begins, so buyers know what they are buying before committing to full diligence spend, and so the QoE team starts from a reconciled base rather than a data room.

AddBack analyzes complete general ledger data and returns normalized EBITDA and financial findings in hours. A traditional QoE process takes three to six weeks and delivers a report with professional sign-off.

AddBack is built for the decision that comes before that engagement: whether a target deserves full diligence spend at all. It reads every entry rather than sampling, reconciles the ledger to the trial balance, and produces an adjustment bridge with entry-level support. It does not issue an opinion a lender can rely on, and it does not negotiate adjustments with the other side.

Most buyers use both. AddBack to screen and prepare, a QoE firm to confirm and sign. Screen first. Spend second.

A Quality of Earnings report for a lower middle market transaction typically costs between $30,000 and $75,000. Smaller deals with clean books can come in between $15,000 and $30,000. Larger or more complex transactions, including multi-entity structures and cross-border deals, commonly exceed $100,000.

Pricing drivers, roughly in order of impact:

• Data condition. Cash-basis books, unreconciled accounts, and missing periods add hours before analysis begins.
• Number of entities. Each legal entity multiplies reconciliation work.
• Periods covered. Three years is standard, and each additional year adds cost.
• Scope. Whether working capital, proof of cash, and debt-like items are included or priced separately.
• Provider tier. National accounting firms price above regional and boutique providers.
• Timeline. Compressed deadlines carry a premium.

Most providers quote a range rather than a fixed fee, and the range widens when they have not yet seen the data.

Request general ledger and trial balance data early, and analyze it before committing to a full engagement. Transaction-level data reveals whether reported EBITDA is supportable well before a QoE firm is involved.

An effective screen answers six questions:

• Does the ledger foot to the trial balance and up to the financial statements
• Do monthly revenue and margin trends look organic, or smoothed
• Are the seller’s proposed add-backs traceable to actual entries
• Is customer concentration disqualifying on its own
• Are there related-party or unusual entries that require explanation
• Is working capital seasonality going to change the price mechanism

Sellers sometimes resist releasing ledger data this early. The resistance is itself information. Reasonable buyers make the request under NDA as a condition of proceeding, and a seller with clean books rarely objects.

The highest-value red flags are revenue recognized ahead of delivery, non-recurring expenses that recur every year, gross margin decline masked by product mix, deteriorating receivable aging, and undocumented owner add-backs.

The working list:

• Revenue spikes concentrated in the final month or quarter before sale
• Add-backs presented without invoices, contracts, or payroll support
• Related-party transactions on non-market terms
• Unrecorded liabilities: accrued vacation, deferred revenue, unpaid payroll taxes
• Ledger totals that do not foot to the financial statements as presented
• Deferred maintenance or capital expenditure suppressing current-period cost
• Inventory or reserve levels inconsistent with historical practice
• Customer concentration paired with expiring or unwritten contracts
• Accounting system changes mid-period, or missing months
• Prior periods restated without clear explanation

Individually most of these have innocent explanations. Clustered, they usually mean reported EBITDA will not survive scrutiny.