How to compare performance across acquired companies
By Derwin Lucas · · Revised
To compare performance across acquired companies, first establish that every company is measuring the same thing. A comparison holds only when each brand counted the same event, over the same population and period, with evidence behind its figure. When those conditions aren’t met, the gap between two brands can be a gap in definitions rather than in performance, and the report alone won’t tell you which.
A platform built by acquisition might run one brand on a dedicated CRM, another on a general-purpose one and a third on monthly spreadsheets. Each reports a close rate, a gross margin and brand EBITDA, and the labels match. This guide sets out four requirements a comparison has to meet, in the order they have to be met, and the mistakes that break each one. All figures are illustrative.
Four requirements for portfolio KPI comparability
Exhibit 1. Four requirements for portfolio KPI comparability
| Requirement | The question it answers | It fails when |
|---|---|---|
| 1. Defined | Does every brand count the same event, divided by the same thing, in the same month? | Two brands report “close rate” for different events |
| 2. Scoped | Which brands, periods, records and exclusions sit inside each figure? | A total covering 22 of 25 brands is read as covering all 25 |
| 3. Traced | Can each figure be traced to the records behind it? | A brand submits a rate without the counts behind it |
| 4. Ruled on | Does the evidence support this particular comparison, and is the reason kept with the answer? | A ranking is published from figures nobody has reconciled |
Each requirement rests on the one before it. A figure cannot be scoped until it is defined, because its scope is the population the definition applies to. It cannot be traced until it is scoped, because there is no fixed set of records to trace it to. And no comparison can be ruled on until the first three are known.
Defined: one meaning per measure
A definition states the event that counts, what it is divided by, and the month an event belongs to. Without one, the same records support several correct answers.
Exhibit 2. One brand’s month, four close rates
| What counts as the sale | Divided by | Close rate |
|---|---|---|
| Gross sales: 10 | 30 demonstrations | 33% |
| Net sales, after cancellations: 8 | 30 demonstrations | 27% |
| Jobs in production: 6 | 30 demonstrations | 20% |
| Jobs in production: 6 | 45 appointments issued | 13% |
If one brand reports the first figure and another the last, the second appears to close at less than half the rate of the first. Their selling doesn’t have to differ at all. See the four close rates
Financial measures behave the same way. Where financing fees, installation labor and sales commissions are booked can put one P&L’s gross margin anywhere from 29.9% to 62.0%. In an illustrative 25-brand portfolio, eight brands that each leave out about $150,000 of shared costs report $1.2 million more brand EBITDA between them than they would on the portfolio’s basis. How to make brand EBITDA comparable
The mistake that breaks this requirement is treating a shared label as a shared definition, and one of its common forms is a reporting template. A template standardizes the boxes; it doesn’t standardize what goes in them. Each brand’s departure from the portfolio definition should be recorded, measure by measure. You don’t need a portfolio definition for every metric, only for the measures behind capital allocation, covenant compliance or valuation.
Scoped: one population per figure
A total that covers 22 of 25 brands is a different total from one that covers all 25, and it should say so and name the three that are missing. A partial year set against a full one, or a newly acquired brand counted for three months against brands counted for twelve, changes the answer before any performance enters it. See the reporting coverage case
The mistake that breaks this requirement is filling the gap. An estimate for a missing brand makes a total look complete when it isn’t, and from that point nobody reading the total can tell which part of it was measured. Missing brands are stated, not estimated.
Scope also covers what a figure leaves out within a brand: which work types, which records and which exclusions. Two close rates on the same definition still differ if one brand excludes financed jobs and the other doesn’t.
Traced: one trail per figure
Tracing is the link between a figure and the records behind it, and how far it reaches depends on what a brand can provide. Record-level data lets you recount a brand under the portfolio definition. A monthly template that reports totals supports less: you can see what the brand submitted, but you can’t recount it. A brand that submits a rate without the counts behind it can’t be restated at all.
None of these is a reason to exclude a brand from reporting. Each is a reason to state what its figure rests on. A figure without a trail can still be reported; what it can support in a comparison is decided under the fourth requirement.
Ruled on: one decision per comparison
The last requirement is a decision, made comparison by comparison, about whether the evidence supports it. There are three outcomes. Where a brand’s records allow, restate its results under the portfolio definition and compare them. Where the records support comparison only with a stated caveat, such as partial coverage or aggregate evidence, compare with that limitation attached. Where they can’t support the portfolio definition, report the number without comparing it, and say why.
The mistakes that break this requirement are comparisons made by default: ranking brands on figures nobody has reconciled, which rewards the brand that measures most generously, and reading revenue levels as performance, which tells you how large a brand is, not how well it performs.
In Goverisyn, every figure carries an evidence state: Substantiated, Qualified, Limited or Not Available. Whether it is compared is a separate ruling. What each state means
Comparability is established in order: defined, scoped, traced, then ruled on.
Cross-brand performance measurement without common systems
None of the four requirements depends on moving brands onto one CRM or ERP. Definitions can be agreed, scopes stated, figures traced and comparisons ruled on while each brand keeps its own systems of record, working from connections, exports, extracts or the monthly submissions it already sends. Why common systems aren’t a prerequisite
Start with one decision measure
Pick the measure your board or sponsor already relies on, often close rate, gross margin or brand EBITDA, and take it through the four requirements across a few brands before extending it. Write down how each brand calculates it, state what each figure covers, note what each rests on, and then decide which comparisons hold.
Goverisyn’s Portfolio Measurement Review does this across selected brands, after a free 45-minute Portfolio Review that needs no data in advance. For the same approach applied to financial results, over time and across trades, see EBITDA comparability across portfolio companies, Definition drift in portfolio reporting and Home-services portfolio reporting across trades.
More guides on comparing brands
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. Published . Revised . About Goverisyn