Quality of Earnings: What PE Sponsors Actually Look For (And What Kills Deals)

A quality-of-earnings report is not an audit. It is a forensic investigation into whether the earnings you are presenting are sustainable, defensible, and purchasable at the multiple you are asking for. The distinction matters, because the most expensive mistake a seller makes is treating QoE as a check-the-box exercise rather than the document that will, more than any other, determine final purchase price.

Having sat on both sides of these processes, here is what sponsors actually test, the red flags that kill deals, and what a finance function should be doing six to twelve months before a process begins.

What a QoE actually tests

A competent QoE is structured around four questions, asked in increasing order of risk.

First: are the numbers real? This is the accounting-integrity layer. The provider will reconcile reported revenue to cash, to signed contracts, and to shipping or delivery evidence. They will test cutoff — whether revenue booked in December actually belongs to December. They will scrub AR aging for customers that no longer exist and inventory for slow-moving or obsolete stock being carried at cost. This layer usually takes two to three weeks and, if the books are clean, produces minor adjustments.

Second: what is the durable earnings base? This is where adjustments get contested. The provider will rebuild trailing twelve-month EBITDA from source data, strip out non-recurring items (both above and below the line management disclosed), and produce an "adjusted EBITDA" number that almost always differs from management's. The delta — the haircut — is the negotiation.

Third: is that earnings base sustainable? Here the analysis moves from backward-looking to forward-looking. Customer concentration, retention cohorts, pricing trends, gross margin by segment, and competitive positioning all feed a view on whether the LTM number is a reasonable proxy for the go-forward number. This is where deals die quietly — not because the numbers were wrong, but because they were unrepeatable.

Fourth: does management know their business? This is the softest layer and the one sellers underestimate most. Across three or four working sessions, the diligence team is forming a view on whether the CFO can explain month-over-month variances without looking at a spreadsheet, whether the commercial leader knows the top ten customers by name and contract terms, and whether the operations lead can articulate why gross margin moved 140 basis points between Q2 and Q3. A finance function that cannot answer these questions fluently will produce a report with a lot of hedging language, and that hedging translates directly into price.

The red flags that kill deals

Not all diligence findings are created equal. Some generate purchase price adjustments; some cause the sponsor to walk. The difference is usually about pattern, not magnitude.

Customer concentration above 30%

A single customer representing more than 30% of revenue is a structural problem for most sponsors, and one representing more than 50% is a deal-breaker for many credit committees regardless of how well the business performs. The issue is not just the loss risk — it is the pricing power dynamic. A concentrated customer sets the price, dictates payment terms, and has optionality on renewal that the seller does not. The earnings that concentration produces are real today and rented tomorrow.

If you are running a process with material concentration, get ahead of it: segment the contract (term length, renewal history, share-of-wallet trajectory), document the customer relationship depth (multiple stakeholders, embedded integration, switching cost), and expect the multiple to be discounted regardless. Sponsors who will pay through concentration are rare and they reward preparation.

Channel or platform dependency

A business that generates 70% of its traffic from Google or 60% of its revenue through Amazon has a different risk profile than a diversified business with the same total revenue. Channel dependency is a version of customer concentration — the "customer" is the platform — and buyers discount accordingly. When Google changes its algorithm or Amazon changes its terms, the business does not get a vote.

Defensible positioning requires a clear view of the unit economics by channel, the CAC-to-LTV ratio on organic versus paid, and credible plans to diversify. Plans without execution history are rhetoric. Execution history without a sustainable trajectory is noise.

One-time revenue dressed as recurring

The fastest way to lose credibility in a QoE is to present ARR that includes non-recurring items. Implementation fees, one-off professional services, and variable consumption overages are not ARR. Buyers have learned to check this — they will ask for an ARR bridge from the prior period, and they will cross-check it against invoice data. A bridge that does not tie will be rebuilt by the provider, and the rebuilt version will be materially lower than what was presented.

This is not a hard problem to avoid. Separate recurring from non-recurring at the ledger level, not just in a waterfall slide. Buyers do not punish businesses that have meaningful non-recurring revenue; they punish businesses that try to classify it as something it is not.

Aggressive capitalization or timing

Capitalization of software development costs is legitimate and often under-utilized. But there is a line between applying ASC 350-40 consistently and capitalizing costs that belong in OpEx. A buyer's diligence team will test the capitalization policy, walk through the specific projects capitalized in the last 24 months, and look at what percentage of the engineering team's time is being capitalized. A policy that suddenly shifts six months before a process — or that captures an unusually high percentage of costs — will be reversed, and the resulting EBITDA restatement is often the single largest haircut in a QoE report.

The same logic applies to revenue timing. Pulling Q1 bookings into Q4 to hit a number is detectable. The cutoff test will find it, and the loss of credibility will cost more than the revenue gained.

Working capital inconsistency

A finance function that cannot explain why DSO moved from 45 to 62 days over the last year has a credibility problem. A business where AP has quietly stretched from 30 to 75 days has a cash flow problem masquerading as a margin improvement. Both will show up in the working capital analysis, and both will show up in the negotiation over the peg.

Revenue quality: the single most important concept

If a sponsor leaves diligence with only one number in mind, it is the revenue quality score — an informal composite of four factors.

1. Recurring versus non-recurring mix. High-quality revenue has a high recurring percentage with documented contracts, renewal behavior, and visible retention.
2. Gross retention and net retention. Gross retention below 85% in a subscription business is a structural concern. Net retention below 100% means the business is losing ground before new sales are counted.
3. Customer acquisition economics. CAC payback period, the ratio of CAC to LTV, and how these have trended over 24 months. A business where payback has extended from nine to eighteen months is growing more expensively — even if revenue is still growing.
4. Concentration and dependency. Customer, channel, and geographic concentration — each measured and each stress-tested.

A business that scores well on all four commands a premium. A business that is weak on any one of them can still transact, but at a discount that reflects the risk.

EBITDA adjustments that survive versus get haircut

We covered the anatomy of a defensible add-back in a prior post. The short version:

• Survive: Transaction costs with invoices. Documented owner comp above market. Concluded legal matters with settlement documents. One-time customer concessions with supporting detail. Discontinued lines with operational evidence.
• Get haircut: "Non-recurring" marketing and rebranding. System implementations at a scale the business should reasonably expect to continue. Growth investment add-backs for hires still on payroll. Management bonuses tied to the transaction without structure.

The pattern: anything that requires the buyer to believe something will not happen in the future is haircut. Anything backed by documentation that something specific already happened and has concluded tends to survive.

Preparing the finance function six to twelve months before a process

By the time a banker is engaged, most of the value a finance function can create has already been created or missed. The work that matters happens in the six to twelve months before a bank is hired.

Months 12 to 9 before process:

• Revenue recognition cleanup, with at least two quarters of clean reporting under the revised methodology
• Capitalization policy review and consistent application, supported by time-tracking data
• Chart of accounts rationalization, so revenue by segment and margin by segment are cleanly reportable
• Month-end close discipline — if close is taking longer than ten business days, it needs attention before diligence, not during

Months 9 to 6 before process:

• Customer cohort analysis and retention reporting — gross and net retention by cohort, by segment, by price tier
• Add-back schedule construction, built as a running file with supporting documentation collected contemporaneously
• 24-month working capital analysis with cycle-adjusted view
• CFO readiness — can the CFO walk a buyer through twelve months of variance fluently? If not, this is the time to fix it

Months 6 to 3 before process:

• Sell-side QoE (commissioned by the seller) to surface issues before a buyer finds them
• Data room preparation — organized by topic, version-controlled, with a clear index
• Management presentation build, tested in front of an independent reviewer
• Gap analysis on anything the sell-side QoE flagged, with remediation plans

Months 3 to 0:

• Final data room review
• Run-rate and pro forma adjustments locked, with supporting evidence
• Management interview preparation — key questions, practiced answers, alignment across the team

The businesses that run clean processes — the ones where multiple bidders hold price and the buyer's final LOI closely matches the seller's expectations — are almost always the ones that started this preparation a year in advance.

The meta-point

A QoE is the most expensive document in a sale process, and it is the one where preparation pays the highest return. The businesses that treat it as a finance exercise underperform; the businesses that treat it as a credibility exercise — about the finance team, the data, and the narrative — capture more value.

We build exit readiness programs specifically around this preparation window, and our strategic finance advisory work focuses on making the finance function itself diligence-ready as part of ordinary operations. If you are inside twelve months of a process — and especially if you are not sure whether you are — that is the conversation worth having now.