Definition drift in portfolio reporting
By Derwin Lucas ·
Definition drift in portfolio reporting is what happens when the meaning of a measure changes over time while its name stays the same. A brand switches CRMs, a controller reclassifies a cost, a template gains a column, and the figure carries on under the old label. The trend line then shows a change that isn’t in the business, or hides one that is.
Drift is harder to see than a difference between brands, because there is no second brand to compare against. The evidence is the history of the definition itself. This guide covers where drift comes from, what it looks like in the numbers, and how to detect and handle it. All figures are illustrative.
Drift versus inconsistency
Inconsistency is two brands measuring differently at the same time. Drift is one brand measuring differently over time.
Both are portfolio reporting inconsistencies, and both make a number mean less than its label suggests. Inconsistency distorts comparisons between brands in the same month. Drift distorts comparisons within a brand, across months: growth rates, year-over-year changes, trailing twelve months and anything measured against plan. A portfolio can fix the first and still suffer from the second, because a definition agreed once does not stay agreed on its own.
Where drift comes from
Drift often arrives with an ordinary operational change, not with a decision anyone recorded as a change of definition.
- A new system. A brand moves to a different CRM or accounting package, and the new system counts appointments, sales or cancellations by its own rules.
- New people. A new controller books commissions or financing fees in a different place from the last one.
- A policy change. A brand changes its cancellation window or the point at which a contract counts as sold.
- An acquisition. An add-on joins the portfolio reporting under the definitions it brought with it.
- A template change. A reporting template adds, renames or drops a column, and brands fill the new version in different ways.
- A portfolio decision. The platform itself starts allocating shared-service costs, or changes how it does.
None of these is a mistake. Each becomes a problem only when the figure carries on as if nothing changed.
What drift looks like in the numbers
Drift shows up as a step change at a month boundary that no operating event explains.
- Close rate. In one illustrative brand’s month, 10 of 30 demonstrations became gross sales and 8 survived cancellations. A brand that switches from counting gross sales to net sales reports its close rate falling from 33% to 27%, with no change in how it sells. See the four close rates →
- Brand EBITDA. A brand reporting EBITDA of $1,500,000 that starts carrying its $150,000 share of shared-service costs reports $1,350,000: 10% lower, with nothing changed in the business.
- Portfolio totals. Scope drifts too. A monthly total that covers 25 brands one month and 22 of 25 the next falls for reasons that have nothing to do with performance. See the reporting coverage case →
Drift can also hide a real change. A brand whose results are genuinely weakening can report a steady figure if its definition becomes more generous in the same month.
Where drift does the damage
Drift does its damage wherever a figure is measured against a fixed reference over time: plan, prior year, a covenant threshold or a trailing twelve months.
- Plan. A budget built on last year’s definition is missed or beaten by a definition change alone.
- Prior year. Year-over-year growth compares two different calculations if the definition moved in between.
- Covenants. A covenant test that holds the threshold steady while the measure underneath it changes can pass or fail for the wrong reason.
- Diligence. A trailing twelve months that spans a definition change mixes two measures in one figure, and a buyer’s quality-of-earnings team is likely to find the seam.
A trend is only as stable as the definition underneath it.
How to detect definition drift
Drift is caught by recording definitions, not by watching the numbers.
1 · Record each definition with the date it took effect
For each measure that matters, write down how each brand calculates it and when that calculation began.
A definition without a start date can’t tell you whether last year’s figure was built the same way as this year’s.
2 · Treat a change in a submission as a question
When a brand’s submission changes form, ask what changed in the definition before accepting the figure.
A new system, a new column, a new person submitting, or counts that stop arriving with a rate are all reasons to ask. Why a template alone doesn’t hold a definition →
3 · Reconcile to source records every period
A reconciliation that tied out last month and doesn’t this month is often the first sign of drift.
Figures that are traced to the records behind them each period carry their own history. Figures that are only submitted as totals don’t.
4 · Screen for unexplained step changes
Look for jumps at month boundaries that line up with a system, staffing or policy change.
A step change is a reason to investigate, not proof of drift. Only the definition’s history can confirm it.
How to handle a definition change
Record the change, restate history where the records allow, and mark the break where they don’t.
Give the new definition an effective date and keep earlier figures under the definition they were built on. Where a brand’s records allow, restate prior periods under the new definition so the trend compares like with like. Where they don’t, show the series as broken at the change rather than splicing two definitions into one line. Missing history is stated as missing, never estimated.
Decide who may approve a change before one arrives. The CFO should approve changes to the measures the board and sponsor rely on, brand controllers should raise a change before it takes effect, and FP&A should record it.
Figures that have already gone to a board or lender need the same care. In Goverisyn, a Portfolio Measurement Statement is versioned and shows what changed since the last statement, so a reader can see when a figure’s basis moved. What a Portfolio Measurement Statement is →
Where to start
Start with the one measure your board watches most closely: write down how each brand calculates it today, and since when.
That record alone will show where the trend can be trusted and where it can’t. Goverisyn’s Portfolio Measurement Review does this across selected brands. The first step is a free 60-minute Portfolio Review, which needs no data in advance. For differences between brands rather than over time, see How to compare performance across acquired companies and EBITDA comparability across portfolio companies.
Start with a Portfolio Review.
60 minutes. No data needed in advance.
By Derwin Lucas, founder of Goverisyn, with nearly three decades in management and information systems in home services. Published . About Goverisyn →