Home-services KPI and EBITDA glossary
By Derwin Lucas · Updated
The home services KPIs and EBITDA terms a multi-brand roofing, windows, bath or HVAC platform reports on, from backlog and customer deposits to covenant EBITDA and close rate. Each entry gives the definition, then where brands that share the term tend to count it differently. That second part is why two brands’ figures can share a name and still not compare.
Revenue, backlog and cash
Revenue in home services can be counted at the sale, at completion or at collection, and brands do not always say which. These terms separate the three, and the cash around them.
Reported revenue
Revenue for the month as the brand submits it to the platform, on the brand’s own recognition policy.
Where brands differ. Under ASC 606, installation on a homeowner’s property is generally recognized over time. Because most jobs finish within days, many brands book at completion as a practical proxy, and some acquired brands still book at the sale or at collection. Two brands with the same reported revenue can be describing different events.
Contracted revenue
The value of work sold in the month: contracts signed, before anything is installed.
Where brands differ. Whether cancelled contracts are removed, and when, differs by brand. Contracted revenue is a sales measure and is never added to reported revenue.
Organic growth
Growth from brands owned in both periods, excluding the effect of acquisitions.
Where brands differ. When an acquired brand enters the base, and whether its prior-year figure is pro forma, change the answer. More in the guide
Insurance and retail mix
The share of revenue from jobs paid through homeowners’ insurance claims, against jobs the homeowner pays for directly.
Where brands differ. Central in roofing and restoration. Brands classify partly insured jobs differently and recognize claim supplements at different points, so close rates and margins are compared within a channel, not across it.
Backlog
Work sold and not yet completed, at contract value, measured at a date.
Where brands differ. Some brands carry cancelled or stalled jobs in backlog for months; others remove them at once. A roll-forward (opening backlog, plus contracted, plus or minus change orders, less cancellations, less completed, equals closing) shows whether the figure closes.
Backlog coverage
Backlog divided by average monthly completed work, in months of work.
Where brands differ. The averaging window matters: one month, three months or twelve give different answers in a seasonal trade. The window belongs in the definition.
Customer deposits
Cash received from homeowners before the work is performed. Under ASC 606 it is a contract liability, reduced as the work is performed; whether it is refundable depends on the contract and the rescission period.
Where brands differ. Some brands credit deposits to revenue on receipt, or net them against receivables across the book; netting is appropriate only within a single contract.
Receivables and days sales outstanding
Amounts billed and not yet collected from homeowners, insurers or finance companies; days sales outstanding expresses them in days of revenue.
Where brands differ. The revenue window used for the days varies, as do whether financed jobs count as receivables until the finance company pays and whether an insurer’s depreciation holdback is included.
Revenue-to-cash conversion
The share of reported revenue received in cash over a period, from a roll-forward of work completed to cash collected. Not the lender’s cash conversion ratio, which starts from EBITDA.
Where brands differ. Brands without payment records can’t produce it. Where they can, cash received but not matched to a contract is the figure to watch.
Gross margin and its lines
Gross margin can move with classification as much as with performance. These are the lines that move it most in home services.
Gross margin
Revenue less cost of goods sold, as a share of revenue.
Where brands differ. What sits in cost of goods differs by brand: sales commissions, installation labor, permits, financing fees and warranty can each sit above or below the gross margin line.
Common-basis gross margin
Gross margin restated so every brand classifies the same costs the same way, shown beside the margin each brand reported.
Where brands differ. It is a measurement reclassification for comparability, not a restatement of any brand’s ledger. Brands that submit totals only are shown at reported margin, outside the common-basis comparison, with the reason stated. More in the guide
Sales commissions
Pay to sales representatives for contracts sold.
Where brands differ. Booked in cost of goods by some brands and in sales and marketing by others; an accounting system conversion can move them between the two mid-year. More in the guide
Installation labor
The cost of the crews that install the work, employed or subcontracted.
Where brands differ. Employee crew burden (payroll taxes, benefits, workers’ compensation, vehicles) often sits in overhead, while subcontractor cost sits wholly in cost of goods, so the same job shows different margins by crew type.
Labor efficiency
Installed revenue per installer-day, or billable hours as a share of paid hours.
Where brands differ. Whether subcontractor days are included, and how travel and callback time are treated, decide the figure.
Financing fees
Fees paid to a consumer finance company when a homeowner finances the job, often called dealer fees.
Where brands differ. Some brands deduct them from revenue, some book them in cost of goods and some in operating expenses. Under ASC 606 a dealer fee is commonly treated as a reduction of the transaction price; the brand’s policy and its auditor decide.
Warranty reserve
An accrual for the expected cost of workmanship claims on completed jobs. A warranty the homeowner pays for separately is deferred revenue, not a reserve.
Where brands differ. Rates differ; some brands expense claims only as they are paid, and a late accrual shows up as a restatement.
EBITDA and its adjustments
One set of books supports several defensible EBITDA figures, each on its own definition. The terms below name each one and what separates them.
Reported EBITDA
Earnings before interest, taxes, depreciation and amortization, as the brands submit them, before any adjustment.
Where brands differ. At platform level, whether holding company costs are deducted changes the figure; it should say which.
Holding company costs
Costs carried by the platform rather than any brand: the finance team, leadership, shared systems.
Where brands differ. Some platforms allocate them to brands and some don’t; brand EBITDA compared across platforms often differs only in this.
Add-back
An expense added back to EBITDA because it is nonrecurring, non-cash or outside the business being valued or lent against.
Where brands differ. Each add-back needs four things on record: that it happened, that it is what it says, that the amount is right, and that it is not already excluded elsewhere. More in the guide
Owner compensation add-back
The difference between a former owner’s pay and what the role would cost to fill. If the owner was underpaid, the adjustment reduces EBITDA.
Where brands differ. The cost to fill the role is a judgment. Supported with judgment and unsupported are different findings and are reported apart. More in the guide
Adjusted EBITDA
Reported EBITDA plus the add-backs management proposes and accepts.
Where brands differ. It is management’s figure, on management’s definition. A quality of earnings report’s diligence-adjusted, or normalized, EBITDA is a different figure, and a lender may accept different add-backs again.
Governed Adjusted EBITDA
Reported EBITDA plus only the add-backs a named client reviewer classified as supported nonrecurring, against criteria the client approved.
Where brands differ. It sits beside management’s figure and never overwrites it, and the difference between them is stated by kind of add-back. It is not a quality of earnings finding or a covenant figure. More in the guide
Covenant EBITDA
EBITDA as the credit agreement defines it, for the borrower group and the test period: the figure a leverage covenant is tested on.
Where brands differ. It follows the agreement’s own caps, conditions, exclusions and pro forma rules, so it can sit above or below adjusted EBITDA. More in the guide
Pro forma EBITDA
EBITDA that counts an acquired company’s earnings for months before it was owned, usually from the seller’s statements, and under many credit agreements certain expected synergies.
Where brands differ. The source of each pre-acquisition month should be recorded; a month with no source is stated as a gap, not estimated.
Run-rate savings
Cost savings counted as if already realized for the whole period.
Where brands differ. They are forecasts, not past costs, and credit agreements commonly admit them only with a cap, a time limit and an officer’s certificate. More in the guide
Sales and marketing
The funnel terms brands share by name and define differently. A comparison holds only when each stage is counted the same way.
Lead
A prospective customer recorded by the brand.
Where brands differ. For some brands a lead is any inquiry; for others, only one with a scheduled appointment. Cost per lead moves with the definition, not the marketing.
Appointment
A scheduled meeting with a homeowner to present the work.
Where brands differ. Counted when scheduled, when confirmed, or when issued to a sales representative, depending on the brand.
Demo held
An in-home sales presentation that took place. Also called a sales presentation or a sit.
Where brands differ. Brands differ on what counts as held: both decision-makers present, a full presentation, or any visit. The rule belongs in the brand’s written definition, approved by the platform.
Gross close rate
Contracts signed divided by demos held in the period.
Where brands differ. Some brands count every signed contract, others only those that survive cancellation; some divide by demos held, others by appointments. More in the guide
Cancellation rate
Contracts cancelled as a share of contracts signed.
Where brands differ. The rescission window, and whether credit declines count as cancellations, differ by brand.
Contract retention
The share of contracts signed that survive cancellation and credit decline. Not the same as customer or membership retention.
Where brands differ. The window for counting a cancellation differs, and brands that don’t report cancellations can’t produce it.
Average contract value
Contracted revenue divided by contracts signed.
Where brands differ. It describes a brand’s mix more than its performance; comparing it across trades, or across retail and insurance work, says little.
Financing penetration
Financed contracts as a share of contracts signed, by count or by value.
Where brands differ. Whether declined applications count, and whether contract value is gross or net of dealer fees, vary.
Marketing cost as a share of revenue
Marketing spend divided by revenue for the period.
Where brands differ. The denominator (contracted or completed revenue) varies, as does whether lead-generation staff and canvassing pay count as marketing.
Cost per contract
Marketing spend divided by contracts signed, by channel.
Where brands differ. Whether spend includes the cost of the sales team, and which channel a contract is credited to, varies by brand.
Maintenance agreement retention
The share of maintenance or membership agreements renewed at term, central in HVAC and plumbing.
Where brands differ. Whether a lapsed agreement renewed within a grace period counts, and whether agreements are counted by customer or by unit, differ.
Operations
How the work itself is measured once it is sold.
Jobs completed
Jobs finished in the period, the count behind revenue per job.
Where brands differ. Whether service calls, warranty visits and change orders count as jobs varies, and so does what makes a job complete.
Revenue per job
Completed revenue divided by jobs completed.
Where brands differ. It moves with what counts as a job, so a brand that counts warranty visits shows a lower figure for the same work.
Job cycle time
Days from contract signed to job completed.
Where brands differ. The start event (signing, measure or permit) and whether permit and material waits are counted differ by brand.
Evidence and reporting
The terms Goverisyn uses to say what a figure rests on and what it supports.
Evidence state
What the evidence behind a figure supports: Substantiated, traced to individual records; Qualified, used with a stated limitation; Limited, not supporting the comparison it appears to enable; Not available.
Note. A figure’s evidence state is separate from whether it enters a comparison.
Comparison ruling
Whether a brand’s figure is admitted to a comparison under the measure’s policy, with the reason when it is not.
Note. An open exclusion, a definition that differs or a rate submitted without counts each keeps a figure out, and the reason travels with it.
Coverage
The brands and months a figure includes, stated beside it.
Note. A total that leaves out a brand is not wrong if it says so; a total that fills the gap with an undisclosed estimate is. Goverisyn does not estimate missing figures.
Restatement
A brand’s replacement of figures it has already submitted, with the reason it gave.
Note. The first submission stays on record, so the change and its reason can be traced.
Issued board pack
The record of a month’s figures, definitions, evidence, qualifications and approvals, issued by the client’s Authorized Issuer and fixed from that moment.
Note. A change after issue is a new version; earlier versions stay on record. More in the guide
Authorized Issuer
The person the client designates to issue the board pack, usually the platform CFO.
Note. Issuing attests to review of the record, not to its accuracy, and is the client’s act, never Goverisyn’s.
Start with a Portfolio Review.
45 minutes. No data needed in advance.
By Derwin Lucas, founder of Goverisyn, with a career in management and information systems in home services. Updated . About Goverisyn