A Quality of Earnings analysis is supposed to help a buyer answer a fairly simple question: Are the company’s earnings real, repeatable, and likely to continue after the deal closes?
Getting to that answer is rarely simple.
The finished report may be clean and organized, but the work behind it often involves weeks of sorting through inconsistent financial data, rebuilding trial balances, mapping accounts, investigating unusual transactions, and going back to management for explanations.
That is where the cost begins to climb.
Quality of Earnings work is often described as a largely manual, Excel-heavy process. There is grunt work at the front end, standardized analysis in the middle, and exception investigation at the end.
That is a useful way to understand why these engagements become so expensive. The real challenge is not any one calculation. It is the amount of manual work required to turn messy company data into something an advisor can actually analyze and defend.
The Analysis Cannot Start Until the Data Is Usable
Companies rarely deliver financial information in a clean, consistent package.
Data may come from different accounting systems, business units, locations, or legal entities. One file may be organized by month, another by transaction date, and another by fiscal period. Account names may change from one year to the next. Customer and vendor names may be entered differently across systems. Some schedules may not tie back to the financial statements at all.
Before an advisor can begin analyzing earnings, someone has to sort all of that out.
That often means cleaning exports, removing duplicates, fixing date formats, reconciling trial balances, correcting inconsistent labels, and tracking down missing information. It may also require combining accounting data with payroll, customer, inventory, or operational records.
This work is necessary, but much of it is not the kind of work buyers believe they are paying for. They are paying for financial insight and professional judgment. Instead, a meaningful portion of the engagement can be spent repairing files and making the data usable enough to begin.
Experienced professionals often find themselves doing work that is repetitive, slow, and difficult to review. The hours add up quickly, especially when every new file creates another round of cleanup.
Trial Balance Normalization Creates More Work Than Most People Expect
One of the biggest challenges is normalizing the trial balance.
A company’s chart of accounts is rarely consistent over several years. Accounts are added, renamed, merged, or used differently over time. A business may also have changed accounting systems, acquired another company, or reorganized its reporting structure.
An expense recorded as “Contract Labor” in one period might appear as “Outside Services” in another. Software costs might be included in office expenses one year and separated into their own account the next. Revenue may be broken out by product line in one system but combined into a single category in another.
Those differences make year-over-year comparisons unreliable unless the accounts are mapped into a common structure.
That mapping is often handled manually in Excel. Someone creates a crosswalk showing where each source account belongs in the normalized financial statements. Then the mapping must be reviewed, corrected, and updated as new information is discovered.
The work becomes especially time-consuming when the logic exists only inside one workbook or in the memory of the person who created it. If a reviewer questions an account classification, the team may have to trace through formulas and tabs to understand what happened.
When a new trial balance arrives, some of the same work may have to be repeated. When a mapping changes, prior schedules may need to be refreshed. When several entities are involved, the complexity grows again.
A single account may not seem important, but dozens or hundreds of questionable mappings can materially affect margins, EBITDA adjustments, and trend analysis.
Standardized Analysis Is Only as Good as the Data Underneath It
Once the information has been cleaned and normalized, the middle portion of a Quality of Earnings engagement tends to follow a more recognizable pattern.
Advisors review revenue trends, margins, customer concentration, payroll, working capital, recurring expenses, related-party activity, and potential EBITDA adjustments. They compare periods, look for unusual changes, and try to separate normal business performance from one-time events.
Much of this analysis can be standardized.
The problem is that the underlying data usually is not.
A workbook can calculate revenue growth perfectly and still produce the wrong conclusion if revenue accounts were mapped incorrectly. A margin chart can look polished while combining categories that were treated differently across periods. A detailed schedule can appear precise even though the source files did not reconcile.
This is one reason Quality of Earnings work requires so much review. The team is not only reviewing the analysis. It is also checking the cleanup, mapping, assumptions, formulas, and source data behind it.
The more manual the process is, the more opportunities there are for small errors to work their way into the final report.
The Real Value Comes From Investigating Exceptions
The most important part of the engagement begins when the analysis identifies something that does not look right.
A sudden improvement in gross margin may reflect better pricing or a shift toward more profitable work. It could also be caused by expenses posted in the wrong period, an inventory adjustment, or a change in how costs were classified.
A sharp increase in EBITDA may indicate genuine growth. It may also be tied to deferred spending, lower owner compensation, a one-time customer order, or expenses that were moved below the line.
The spreadsheet can identify the change. It cannot explain the business reason behind it.
That requires investigation.
The advisor may need to trace transactions into the general ledger, review invoices or contracts, compare accounting records with operational systems, interview management, and request additional schedules.
Each answer can lead to another question. An explanation from management may reveal a classification problem. A classification problem may require a revised mapping. The revised mapping may change several schedules that were already completed.
This is where deals begin to slow down.
The investigation itself is valuable. The inefficiency comes from how it is managed.
In many engagements, open questions are spread across spreadsheets, emails, meeting notes, and data request lists. Different team members may be working from different file versions. It may not be clear which questions have been answered, which explanations have been verified, or whether the analysis was updated after new information arrived.
The work is not expensive because advisors are asking too many questions. It is expensive because the process for managing those questions is often fragmented.
Every New File Can Restart Part of the Process
Quality of Earnings work is rarely completed in a straight line.
The seller provides the initial financial package. The advisory team begins its analysis. Questions arise. Management sends revised schedules or additional exports. The team incorporates the new information and discovers that earlier assumptions need to change.
A correction to the chart-of-accounts mapping may affect revenue, margins, operating expenses, and EBITDA adjustments. A new customer file may change the concentration analysis. An updated payroll schedule may require several months of data to be reclassified.
In an Excel-heavy process, each revision carries risk.
Files are copied and renamed. Formulas are extended. Tabs are added. Data is pasted into existing templates. Notes are updated in one place but not another. Team members may continue working from an older version without realizing it.
Even when the underlying issue is relatively small, the effort required to update and re-review the analysis can be significant.
That cycle is a major reason fees climb. The team is not simply performing the analysis once. It may be partially rebuilding and checking it several times as the engagement develops.
Higher Fees Do Not Automatically Produce Greater Confidence
The frustrating part is that a long, expensive engagement does not always leave the buyer completely confident in the result.
The buyer may still want to know where a number came from, how an account was classified, why an adjustment was made, or whether management’s explanation was supported by the data.
Those questions are easier to answer when the process is traceable.
A reviewer should be able to move from a conclusion in the final report back to the supporting analysis, the normalized account, and the original source data. It should also be clear when the information was received, who reviewed it, and whether later changes affected the conclusion.
When that history is spread across spreadsheets and email chains, the final answer may be correct, but it is harder to prove.
Trust does not come from adding more tabs to a workbook. It comes from being able to show how the team reached its conclusion.
A More Repeatable Approach to Quality of Earnings
The answer is not to remove professional judgment from the process. Judgment is the reason experienced advisors are involved in the first place.
The opportunity is to reduce the amount of manual work surrounding that judgment.
A more repeatable Quality of Earnings process would bring data ingestion, trial balance normalization, account mapping, variance analysis, exception tracking, review notes, and workpaper production into one connected workflow.
Source files could be recorded by entity, period, system, and version when they are received. Trial balances could be mapped using documented rules rather than one-off formulas. Reconciliations could happen before the analysis begins. Standard schedules could refresh when corrected data is added.
When an exception is identified, the question, supporting documents, management explanation, reviewer notes, and final resolution could stay connected to the underlying transaction or account.
That would not eliminate the need for investigation. It would make the investigation easier to manage and easier to review.
It would also reduce the amount of time senior professionals spend checking whether the latest spreadsheet includes the latest answer.
Where AI Can Help—and Where It Cannot
AI can be useful in Quality of Earnings work, but only after the process is under control.
It may help classify accounts, organize supporting documents, summarize management responses, identify unusual changes, or flag transactions that deserve closer review.
What it cannot do is make unreliable data reliable.
If account mappings are inconsistent, source files do not reconcile, or business rules have not been documented, AI can produce a confident-sounding analysis that is still wrong.
That creates more risk, not less.
The better approach is to use AI inside a controlled workflow, where its output can be traced, reviewed, and approved. The goal should be to reduce repetitive effort while keeping financial professionals responsible for the conclusions.
AI should help advisors spend less time searching through files and more time understanding what the numbers actually mean.
Quality of Earnings Will Always Require Judgment
Quality of Earnings work is expensive because it combines messy data, detailed financial analysis, repeated revisions, and high-stakes professional judgment.
The judgment is not the problem.
The problem is that too much of the process still depends on manual cleanup, disconnected spreadsheets, and one-off workflows that have to be rebuilt for every engagement.
Accounting firms, CFOs, and transaction advisors have an opportunity to make the process more consistent without turning it into a black box. The repetitive work can be structured. The mapping logic can be preserved. Exceptions can be tracked in one place. Workpapers can be connected directly to their supporting data.
That allows experienced professionals to focus on the part buyers value most: understanding the business, challenging the assumptions, and determining whether the earnings will hold up after the deal closes.
Ayoka develops custom software for complex financial and operational workflows. For firms performing Quality of Earnings work, that can mean building a repeatable process for data ingestion, normalization, analysis, exception management, review, and final workpaper production.
The result is not just a faster engagement. It is a clearer process, a stronger audit trail, and a financial analysis that buyers can trust.
