Somewhere around week two of due diligence, the request arrives: revenue and gross margin by service line, monthly, for the last thirty-six months.
For a lot of owners this is the moment the trouble starts. The business genuinely has two distinct lines. Everyone internally knows one is better than the other. But the accounting system was never built to tell them apart, so the answer has to be assembled by hand in a spreadsheet, from invoice exports and memory, under a deadline, for a buyer who is now going to weigh that reconstruction against numbers pulled straight from the general ledger.
That is a bad position, and the cost of it is not abstract. It shows up as a lower multiple, and the arithmetic behind why is worth walking through carefully, because it is larger than most owners expect.
A Multiple Is a Weighted Average, and Yours Is Getting Dragged Down
Companies do not trade at one multiple. Revenue streams trade at multiples, and the number on your deal is the weighted average of them.
Consider a business with $10M in revenue and $2M in EBITDA. Inside it are two lines. One is recurring, contracted, high margin, and predictable, producing $1.2M of that EBITDA. The other is project-based, lumpy, lower margin, and dependent on winning new work each year, producing the other $800K.
Priced separately, those are different assets. The recurring line might support six times EBITDA. The project line might support three and a half. Sum of the parts: $7.2M plus $2.8M, or roughly $10M of enterprise value.
Now price it as one undifferentiated business. The buyer sees $2M of blended EBITDA with meaningful lumpiness in it and no way to isolate where the durability lives. They apply something in the neighborhood of four times. That is $8M.
Same company. Same cash flows. Same customers. A $2M difference, and the only variable that moved was whether the general ledger could separate the two lines.
Why buyers default to the conservative number
The instinct is to argue this is unfair, since the recurring revenue exists whether or not the accounting system tracks it. That argument loses, and understanding why it loses is useful.
A buyer underwrites what can be verified. When you assert that 60% of EBITDA comes from a recurring line and the ledger cannot corroborate it, the assertion is not treated as false. It is treated as unproven, and unproven claims get priced at the conservative end because the buyer is the one carrying the risk of being wrong.
There is a second dynamic on top of that. A reconstruction you assemble during diligence is an argument you are making about your own business, prepared by you, under pressure, in a negotiation. Contemporaneous ledger data is evidence. Diligence teams are trained to weight those differently, and they should.
In the IBBA and M&A Source Market Pulse survey, conducted with Pepperdine’s Private Capital Markets Project, intermediaries named unrealistic seller value expectations the single most common reason listed businesses fail to sell, ahead of both buyer and seller preparation.
Source:
IBBA, M&A Source, and Pepperdine Private Capital Markets Project, Market Pulse Survey
Worth sitting with that finding for a second. Some of those sellers were being unrealistic. Others knew exactly what their business was worth and had no way to demonstrate it, which looks identical from the buyer’s side of the table.
Your Chart of Accounts Is a Data Model
Most owners think of the chart of accounts as a filing system for the CPA. It is closer to a database schema. It determines which questions the business is structurally capable of answering, forever.
That last word is the important one. If a transaction is recorded without a segment identifier, the information is not hidden somewhere waiting to be found. It was never captured. You can sometimes infer it later from surrounding evidence, which is slow and imperfect, but you cannot query a dimension that does not exist.
Which means the design decision made casually in year two, usually by whoever set up the accounting file, quietly determines what the business can prove in year eight when it matters most.
The mistake almost everyone makes when they try to fix it
The obvious fix is to create separate accounts for each line. Revenue Line A, Revenue Line B, Direct Labor Line A, Direct Labor Line B, and onward.
This works at two lines and collapses immediately after. The account count multiplies rather than adds. Forty natural accounts across three service lines becomes a hundred and twenty. Add two locations and it is two hundred and forty. Reconciliation becomes miserable, misposting becomes routine, and the financial statements turn into something no outside reader can follow. Businesses that go down this road also tend to collide with their accounting platform’s account limits sooner than they expect.
The correct architecture separates two different ideas that duplicated accounts jam together. Natural accounts describe what kind of transaction occurred: revenue, direct labor, materials, subcontractors, rent, insurance. Dimensions describe which part of the business it belongs to: service line, location, department.
Keep the natural accounts short and stable. Put the segment on a dimension. Now three service lines and forty accounts is still forty accounts, and you can report on any combination.
What that looks like in the systems you actually use
In QuickBooks Online, the dimension is Classes, available on Plus and Advanced only. Simple Start and Essentials do not have it, and no workaround changes that.
Turn it on under the gear icon, then Account and Settings, then Advanced, then Categories. Two settings there matter more than anything else in this article.
First, in the Assign classes dropdown, select “One to each row in transaction” rather than one per transaction. This is the difference between a system that can handle a single invoice covering both service lines and one that forces you to pick one and misstate the other. Most invoices in a two-line business are mixed. Getting this wrong at setup poisons the data quietly.
Second, enable the warning for transactions entered without a class. Without it, unclassified entries accumulate silently and surface later as a “Not Specified” column on your Profit and Loss by Class, which is the report a buyer will eventually read.
Know your ceiling before you design. If you intend to track service line and location and department at the same time, the Plus caps arrive faster than expected, since sub-classes count against the same limit.
QuickBooks Online Plus caps you at 40 classes and locations combined and 250 chart of accounts entries. Advanced removes both limits. Class tracking is unavailable entirely on Simple Start and Essentials, so the plan you chose years ago may already be deciding what you can report.
Source:
Intuit, QuickBooks Online usage limits documentation
Advanced also includes a Reclassify Transactions tool for batch corrections, which is genuinely useful during a cleanup and absent on Plus.
Xero handles this through Tracking Categories, with a harder constraint. You get four defined but only two active at once, and Xero recommends staying under about a hundred options per category for report performance. Two active dimensions means choosing deliberately, usually service line plus one of location or customer type. A single transaction line also cannot be split across tracking options in Xero, so the line-level detail on source documents has to carry more weight.
Businesses genuinely needing three or more simultaneous dimensions have outgrown both. That is the point at which Sage Intacct or NetSuite stop being overkill.
Where Segment Reporting Actually Breaks: Allocation
Tagging revenue is easy. Tagging direct costs takes discipline. Allocating shared overhead is where most implementations die, and there is a widely repeated piece of advice here that is wrong for valuation purposes.
Costs sort into three groups. Direct costs attach to one line without argument: the technician on that job, the materials on that order, the subcontractor on that project. These get tagged at the source document, on the bill or the timesheet or the purchase order, not corrected later.
Semi-direct costs attach to more than one line but have a defensible driver. A project manager splitting time across both lines gets allocated by hours. A delivery vehicle serving two lines gets allocated by trips or mileage. These require a stated methodology and consistent application.
Shared overhead genuinely serves the whole company: the CFO, the office lease, general liability insurance, the accounting system, the owner’s compensation.
Stop short of full allocation
Standard advice says allocate everything so each line carries its fair share and you can see true segment profitability. For internal management that is defensible. For a transaction it is a mistake, and here is why.
Every allocation is a methodology choice, and every methodology choice is arguable. Allocate overhead by revenue and the high-revenue line looks worse. Allocate by headcount and the labor-heavy line looks worse. Allocate by transaction volume and you get a third answer. When you hand a buyer fully-loaded segment income statements, you have handed them a set of assumptions to attack, and any line whose profitability depends on your chosen allocation basis becomes negotiable.
The format that holds up is the one transaction advisory teams and private equity firms use internally. Revenue by line. Direct costs by line. Contribution margin by line. Then unallocated corporate overhead as a single block below, not pushed into the segments at all.
That presentation is harder to argue with because everything above the contribution margin line traces to a source document. Nobody can dispute that a specific technician’s wages belong to a specific job. They can dispute whether 60% of the CFO’s salary belongs to the recurring line, so you decline to make that claim.
Keep a fully allocated view for your own decisions if it helps you run the business. Lead with contribution margin when someone is valuing it.
The Pass-Through Problem
One version of this issue deserves separate treatment because it distorts everything above it, and it is common in agencies, contracting, and distribution.
Some service lines carry substantial pass-through costs. A marketing agency buying media on behalf of clients. A general contractor carrying subcontractor billings. A distributor with drop-ship arrangements. In each case gross revenue includes money that flows straight through with little or no margin attached.
An agency billing $10M gross with $6M of client media spend running through it has $4M of net revenue. Its real gross margin lives on that $4M. Lump that line together with a pure services line under one revenue account and the blended margin looks weak, the services line’s actual health disappears into the average, and any margin trend gets driven by fluctuations in media spend that have nothing to do with operating performance.
Recall that diligence teams decompose margin movement into price, volume, mix, and cost. If pass-through volume is swinging your reported margin and you cannot isolate it, that decomposition produces conclusions about your business that are simply wrong, and you will be arguing against your own numbers.
The fix belongs at the account level, not the dimension level. Pass-through revenue and its matching cost want separate natural accounts so net revenue is calculable directly from the general ledger without a schedule.
What Three Different Buyers Do With Segment Data
Segment reporting matters differently depending on who is across the table, which is worth knowing because it changes what you should emphasize.
A strategic acquirer often wants one of your lines rather than both. They have a business that the recurring line would slot into, and the project line is a distraction they would need to divest or run down. Without segment financials they cannot evaluate the carve-out, cannot model what they are actually buying, and frequently cannot get it approved internally. Some walk at that point, not because the asset is bad but because the analysis is impossible.
A private equity buyer is asking which line scales. They are underwriting a growth thesis and need to know whether returns come from expanding the recurring line, professionalizing the project line, or acquiring similar businesses. All three require segment-level unit economics.
A lender is sizing debt against durable cash flow. Recurring contracted revenue supports more leverage than project revenue, and lenders will attribute the entire EBITDA base to the weaker category when the split is unproven. That directly constrains how much debt the buyer can raise against your business, which constrains what they can pay for it.
There is also the growth story to consider. Owners frequently tell buyers the business is shifting toward the better line. That statement is worth real money when thirty-six months of segment data show the mix moving quarter by quarter. Without the data it is a claim about the future made by the person selling the asset, and it gets weighted accordingly.
Can You Fix It Retroactively
Partly, and the honest answer matters because it determines how much runway you need.
Revenue is usually the most recoverable. If invoices were issued with distinguishable line items or service codes, you can export invoice-level detail, map each item to a service line, and rebuild monthly revenue by segment. In QuickBooks Online, the Sales by Product/Service Detail report gives you the raw material, and a mapping table joined with SUMIFS in Excel reconstructs the history reasonably well. This works when your item list was maintained with any discipline and fails when everything was billed against a generic “Consulting Services” item.
Direct labor is recoverable when timesheets exist with job or client coding. Materials and subcontractor costs are often recoverable from bill detail if the vendor or purchase order identifies the job.
What does not come back is anything recorded only as a summary journal entry, and anything where the underlying document never carried the distinguishing information. If a month of revenue was booked as a single entry from a spreadsheet, that month cannot be split by anything other than an assumption.
Two practical notes. A rebuild of twenty-four months is a project measured in weeks, not an afternoon, and it needs someone who can defend the mapping decisions afterward. And buyers do discount reconstructed data relative to contemporaneous ledger data, which means the version you build in year six is worth less than the version the system captured in year four. It is still far better than nothing, and it is dramatically better done before diligence than during it.
Building It Properly
The setup work is not large. It is the ongoing discipline that decides whether the data is worth anything.
Design the natural accounts first and keep them lean. Sixty to eighty accounts covers most businesses in this revenue range. Group them conventionally so an outside reader can navigate without a legend: 4000s for revenue, 5000s for direct costs, 6000s for operating expenses, 7000s for other income and expense. Within revenue, split by recognition character rather than by segment, since recurring versus one-time versus pass-through is a different axis from service line and belongs at the account level where it is visible on the face of the P&L.
Then define the dimension. Two or three service lines, named the way the business actually talks about them, not the way a consultant would. Resist adding a fourth until the first three are being applied cleanly.
Push the tagging as far upstream as you can. Map products and services items to default classes so invoices carry the segment automatically. Set vendor defaults where a supplier only ever serves one line. Every tag applied at entry is a tag nobody has to fix at close.
Then measure the thing that determines whether any of it worked: run Profit and Loss by Class monthly and check the Not Specified column. Under two or three percent of revenue is workable. Above that, the segment data is decorative, and the month-end close needs a step where unclassified transactions get resolved before the books are declared final.
Set a policy for the semi-direct allocations in writing, apply it consistently across periods, and do not change it midstream. A methodology that shifts between years produces trend lines nobody can interpret, including you.
The Bottom Line
Two service lines under one chart of accounts is not a bookkeeping preference. It is a decision to keep your best economics unprovable.
The lines exist regardless. The recurring revenue is real, the margin difference is real, and the mix shift you have been working on is real. None of that gets valued at full weight unless the general ledger can demonstrate it without a spreadsheet built during diligence.
The build costs a weekend of design and a few minutes per week of discipline. The gap between blended pricing and sum-of-the-parts pricing on a business of any size runs to seven figures. That trade is not close, and the only genuine cost is that the benefit arrives years after the work, which is precisely why it keeps not getting done.
Where to Go From Here
Restructuring a chart of accounts is straightforward in principle and easy to get wrong in the details, particularly the allocation policy and the decision about which dimensions the business will actually maintain.
At Frak Finance, chart of accounts design and cleanup is foundational work, whether the goal is running the business on better information or making sure the numbers hold up when someone is valuing them.
Schedule a free consultation andwe’ll look at whether your books can currently prove what your business is actually worth.
