The Month Is a Lie

The previous three articles changed the unit of measurement, the denominator, and who gets credit.

This one changes the time horizon.

Every home improvement operator reviews monthly numbers.

Revenue for the month. Leads for the month. Close rate for the month. Cancel rate for the month. Marketing cost for the month.

The monthly report is the default unit of analysis for almost every operator in the industry. It is how performance is measured, communicated, and evaluated. It is how bonuses are calculated, budgets are justified, and decisions are made.

The monthly report is also one of the least reliable instruments available for understanding how a business is actually performing.

Not because the numbers are wrong. Because a month is not a unit of business performance. It is a unit of time.

Here is the problem with monthly reporting.

A contract signed on the first of the month and a contract signed on the last of the month both appear in the same revenue period. But their outcomes -- whether they survive to become installed, paid jobs -- will not be known for four to eight weeks. By the time those outcomes are visible, they belong to a different month.

The month that produced the revenue and the month that reveals its quality are not the same month.

Monthly reporting closes a period. It does not follow a cohort.

The result is a persistent distortion. A month that looks strong may be drawing on contracts signed in a prior period that happened to survive. A month that looks weak may be producing a cohort of high-quality contracts whose retention won't be visible for weeks. The monthly number tells you what settled. It does not tell you what was built.

The instrument itself is incapable of answering the question most operators are actually asking.

Cohort analysis solves this by changing the unit of measurement.

Instead of asking what happened in October, cohort analysis asks what happened to the contracts signed in October.

The cohort is every contract signed in a defined period. The analysis tracks that cohort forward through time -- through the waterfall stages, through financing, through scheduling, through installation -- until every contract in the cohort has either become installed revenue or exited the system.

The result is not a monthly snapshot. It is a complete picture of what a period of signing actually produced.

That picture looks different from the monthly report. It takes longer to complete. And it reveals things the monthly report will never show.

What cohort analysis reveals that monthly reporting cannot

Lead source quality over time.

A monthly report shows last month's cancel rate by source. Cohort analysis shows how the cancel rate for contracts signed through a given source has changed across six consecutive cohorts.

Those are not the same thing.

A source whose monthly cancel rate looks stable may be masking a trend. If the cohort from four months ago had a 12% cancel rate and the cohort from last month had a 19% cancel rate, something has changed in that source's lead quality. The monthly number averages it. The cohort analysis shows the direction.

Sources deteriorate. Targeting drifts. Competition increases. Shared platforms reduce qualification standards to maintain volume. These changes show up in cohort performance before they show up in monthly metrics. Sometimes months before.

Rep performance over time.

Monthly close rate and retention-adjusted close rate are snapshots. Cohort analysis by rep shows whether a rep's retention performance is improving, declining, or stable across consecutive signing periods.

A rep whose retention-adjusted close rate has declined in four of the last six cohorts is showing a trend that no single month makes visible. That trend is a coaching conversation. Or a compensation conversation. Or a territory conversation. None of those conversations can happen if the measurement stops at the monthly snapshot.

Operational capacity signals.

When pre-install cancellation rates rise in a cohort, the cause is usually not in the contracts themselves. It is in what happened between signature and installation. Scheduling delays. Communication gaps. Extended timelines. Installation backlog.

Cohort analysis connects the cancellation event to the operational conditions that existed when those jobs were moving through the pipeline. Monthly reporting cannot make that connection because by the time the cancellation appears, the month has closed and the operational context has changed.

The vintage question

Cohort analysis eventually produces a vintage comparison.

A vintage is a cohort treated as an investment period. The contracts signed in Q1 are a vintage. Q2 is a different vintage. The comparison between them reveals how the quality of the business's output is changing over time.

PE firms use vintage analysis to evaluate whether an operating company is improving or deteriorating at the level that matters: not revenue, but the quality of revenue being produced.

A business can produce six consecutive months of stable revenue while every successive vintage deteriorates.

The income statement says nothing is wrong. The cohorts are already warning you.

A business whose Q4 vintage shows higher retained revenue per contract than its Q2 vintage is improving. The contracts are surviving at a higher rate. The denominator is growing. The quality of what the operation produces is moving in the right direction.

A business whose recent vintages show declining retention is deteriorating. The monthly revenue may look stable. The vintage analysis shows the foundation eroding before the financial statement reflects it.

This is the early warning function that institutional operators build and most home improvement operators have never seen.

Building cohort analysis

Cohort analysis requires the same three data connections the waterfall requires: the CRM, the financing platform, and the operations system.

The difference is time. The waterfall is a snapshot of a completed period. Cohort analysis is an ongoing tracking function. Contracts signed today become the current cohort. Their outcomes will populate over the next eight to twelve weeks.

This means cohort analysis is never finished. It is always in progress. At any given moment, the most recent cohorts are incomplete. The older cohorts are closed. The comparison between them is where the operational intelligence lives.

Most operators do not have this infrastructure. Not because it requires sophisticated technology. Because it requires connecting systems that were never designed to talk to each other and maintaining the discipline to track outcomes forward in time rather than backward from the monthly close.

The operators who build it gain a visibility advantage that compounds.

Every cohort adds data. Every vintage adds comparison. Every comparison adds resolution. The picture of how the business actually performs -- not what it reports, but what it produces -- becomes clearer with every period.

Monthly reporting will always be necessary.

It closes the period. It settles the numbers. It creates the shared version of reality that every organization needs to function.

But it answers only one question.

The month tells you what happened.

The cohort tells you whether what happened is getting better or worse.

Those are not the same question. And only one of them is useful for governing a business toward a better future.