The Business Inside the Business

The previous article introduced the cohort as the correct unit of measurement.

This one shows what happens when you compare them.

A home improvement business can grow revenue for three consecutive years and be deteriorating.

Not in the sense that something feels wrong. Not in the sense that anyone is concerned. Revenue is up. The team is larger. The pipeline looks healthy. The monthly report looks fine.

But underneath the revenue line, something is changing.

The contracts signed last quarter are surviving at a lower rate than the contracts signed the quarter before. And the quarter before that. The trend has been running for eighteen months. Nobody has seen it because nobody has been measuring the right thing.

The income statement shows a growing business.

The vintage analysis shows a business whose foundation is quietly eroding.

Retention rate by vintage is the instrument that separates these two pictures.

The concept is straightforward. Take every cohort of contracts signed in a defined period -- a quarter, a month, a half-year -- and calculate the percentage that survived to become installed, paid revenue. That survival rate is the retention rate for that vintage.

Now compare retention rates across consecutive vintages.

If Q1 retained 81% of signed contract value, Q2 retained 78%, Q3 retained 74%, and Q4 retained 71%, the business has a retention trend. Not a retention rate. A trend.

The trend is the information.

The individual quarter is just the data point.

Why the trend matters more than the rate

Most operators who track cancel rate manage it as a target. The cancel rate is 18%. The goal is to get it below 15%. Progress is measured against the target, not against prior performance.

This produces a management behavior oriented toward the number rather than the direction.

An operator can hold cancel rate at 18% for twelve months and believe the business is stable. But if the contracts being signed in month twelve are lower quality than the contracts signed in month one -- if the lead mix has shifted, if the sales process has drifted, if the operational execution has degraded -- the 18% is masking a deterioration that will eventually surface as something worse.

Retention rate by vintage forces a different question.

Not: are we at the target?

But: is the quality of what we produce improving or declining?

Those questions lead to different conversations. Different diagnoses. Different interventions.

What causes vintage deterioration

Retention rate by vintage declines for identifiable reasons. The trend is not random. When retention deteriorates across successive vintages, the cause is usually traceable to one of four sources.

Lead mix shift.

As operations scale, lead mix often shifts toward higher-volume, lower-quality sources. The revenue growth that looks like success is being built on a foundation of contracts with lower survival rates. The monthly close doesn't show the shift because volume is increasing. The vintage analysis shows it because retention is declining.

Sales process drift.

Sales floors evolve informally over time. New reps join with different habits. Successful closes get imitated without their underlying discipline. Qualification standards erode in pursuit of volume. The contracts being signed in month eighteen look like the contracts signed in month one on the surface. They do not survive at the same rate.

Operational capacity compression.

As volume increases, operational capacity compresses. Scheduling timelines extend. Pre-install communication deteriorates. Installation quality becomes more variable. The contracts that were signed into a well-functioning operation now enter a strained one. The survival rate reflects the operational conditions, not just the sales process.

Market condition changes.

Interest rates rise. Consumer confidence shifts. Competition intensifies in specific channels. The buyers who were signing and completing six months ago are signing and cancelling today. Not because the operation changed. Because the environment did.

Each of these produces a retention trend. Each requires a different response. None of them are visible in a monthly cancel rate report.

The healthy vintage signature

Institutional operators who build vintage analysis eventually develop a baseline for what healthy retention looks like in their market and vertical.

The specific number varies. Bath remodeling retains differently than roofing. High-ticket windows retain differently than entry-level solar. The baseline is always specific to the operation, the vertical, and the market conditions.

What is not specific is the direction.

A healthy operation shows stable or improving retention rates across consecutive vintages. The quality of what is produced -- not the volume, not the revenue -- holds or improves over time. When it does, scaling that operation compounds. More contracts at the same retention rate produces proportionally more retained revenue.

A deteriorating operation shows declining retention rates across consecutive vintages.

Scaling does not solve the problem. It amplifies it.

More contracts at a declining retention rate accelerates the gap between what is signed and what survives. Most operators discover this only when the revenue gap becomes impossible to ignore. By then, the trend has been running long enough that the intervention is expensive.

The operators who measure first scale differently.

The vintage conversation PE firms have that most operators never do

When a PE firm evaluates a home improvement platform, they build the vintage analysis before they build almost anything else.

Not because they distrust the revenue number. Because the revenue number does not tell them what they need to know.

They need to know whether the business is producing contracts of stable quality or declining quality. Whether the retention trend is positive or negative. Whether the gap between signed contract value and retained revenue is widening or narrowing over time.

Those questions cannot be answered by the income statement. They can only be answered by cohort data.

An operator who can present clean vintage analysis -- three years of quarterly cohorts showing stable or improving retention rates across changing lead mix, volume, and market conditions -- is presenting evidence that the business works. Not just that revenue has grown.

Revenue can grow for reasons that do not reflect operational quality. Favorable market conditions. Competitor failures. Temporary volume spikes. Vintage analysis removes those explanations and shows what the operation itself produces when contracts are tracked to their outcome.

That is a fundamentally different picture than revenue growth.

Building retention rate by vintage

The same infrastructure that produces the waterfall and cohort analysis produces vintage tracking. The CRM provides signature dates and early dropout. The financing platform provides approval and fallout data. The operations system provides installation completion and cancellation events.

The addition is time-based segmentation. Contracts are grouped by signing period and tracked as a unit. The output is a retention rate per cohort that can be compared across periods.

Most operators have the data. What they lack is the segmentation discipline -- the practice of grouping contracts by when they were signed rather than when they closed, and following each group forward to its outcome.

That discipline, consistently applied over six to eight quarters, produces a vintage record. The vintage record becomes one of the most valuable analytical assets the business owns.

Not because it explains the past. Because it predicts the future.

A business with a stable vintage record entering a period of scaling has evidence that the operation can sustain quality at higher volume. A business with a declining vintage record entering the same period has a warning that scaling will amplify the problem rather than solve it.

The vintage does not lie.

Revenue reports what the business collected.

Vintage analysis reports what the business is becoming.

One closes the period.

The other reveals the trajectory.