Net revenue is the score. Retained revenue is the diagnosis of why the score is what it is. Most home improvement operators can tell you their net revenue. Almost none can tell you their retained revenue rate by cohort, by source, or by rep.
Every home improvement operation produces two revenue numbers. The first is what gets reported: net revenue, sold minus cancellations, refunds, and credits. It shows up in the monthly close, it drives bonus calculations, and it tells leadership whether the business hit its target.
The second number almost never gets reported. It is the rate at which signed contract value survives the operational journey from signature to deposited cash, measured by the cohort that signed in a given period. That number is retained revenue. And in home improvement, it is the only number that tells you whether the growth you are reporting is real or borrowed.
Why Net Revenue Is a Lagging Indicator
The problem with net revenue as a management metric is not that it is inaccurate. It is accurate. The problem is that it is a settlement of decisions made weeks or months earlier. By the time a cancellation appears in the monthly close, the job was sold, the deposit was processed, the install was scheduled, and the cancellation was processed through operations and accounting. That sequence takes time. Often four to eight weeks.
This means that a strong net revenue month can coexist with a deteriorating operational structure. The jobs closing well in April may be the product of a February cohort that was already leaking at every stage. The cancellations from that February cohort will not appear in the report until April or May, after the accounting close catches up with the operational reality.
The April report looks like growth. The February cohort retained 77 cents of every dollar sold. The operation booked a full quarter of new jobs against a structure that was leaking 23 percent of its signed value before the accounting system had time to surface it.
By the time the net revenue report reflected the problem, the decisions about March and April hiring, lead spend, and capacity had already been made on the basis of a number that was not yet telling the truth.
The Four Stages Where Retention Leaks
Retained revenue is not lost in one place. It leaks across four distinct stages, each governed by different variables and requiring different interventions. A blended cancellation rate obscures which stage is driving the loss. Measuring cancel rate as a single number produces a single intervention response aimed at the wrong stage.
What it is: Signed contracts that cancel during the rescission or cooling-off window, typically within 72 hours to 7 days of signature. Often driven by buyer's remorse, spousal disagreement not surfaced during the demo, or financing approval failure. What drives it: Overselling, pressure close tactics, financing presented too late in the process, or lead sources that attract lower-intent buyers. This stage is almost entirely a sales execution and lead quality signal.
What it is: Contracts where the financing application is denied or the customer declines the approved terms after seeing the actual payment. Distinct from Stage 1 in that the buyer may have wanted to proceed but the financing structure failed them. What drives it: Financing options not matched to buyer profile, approval rates by lender not tracked against cancel rates, or reps presenting financing without understanding approval probability by product tier.
What it is: Contracts that survive the rescission window and financing approval but cancel before installation is completed. The deposit has typically been collected. The job has been scheduled. The cancel happens between weeks two and eight. What drives it: Install scheduling delays, competitive re-shopping during the wait window, change in household circumstances, or post-sale communication gaps that allow buyer doubt to compound without intervention. This is the most expensive stage on a per-job basis because the acquisition cost is fully sunk.
What it is: Jobs that complete installation but generate partial or full refunds due to quality disputes, warranty claims processed as refunds, or chargeback activity. The revenue appeared in the close but was clawed back in a subsequent period. What drives it: Install quality variance, product representation during the demo that does not match the delivered product, or warranty handling that routes to refunds rather than remediation.
Each stage requires a different fix. Stage 1 is a sales and sourcing problem. Stage 2 is a financing architecture problem. Stage 3 is a post-sale operations problem. Stage 4 is an install quality and warranty handling problem. Treating all four as a single cancel rate metric produces the wrong diagnosis and the wrong intervention at every stage.
How to Calculate retained revenue rate by cohort
Retained revenue is calculated at the cohort level, not the monthly accounting level. A cohort is the set of contracts signed in a defined period, typically a calendar month. The cohort is tracked forward in time until all four leak stages have resolved, which typically requires 90 to 120 days from the signature date.
Three data sources are required that most operations have never connected: the signed contract record from the CRM, the cancellation and refund record from operations or accounting, and the financing approval record from the financing platform. None of these systems were designed to talk to each other. Retained revenue is what becomes visible when they do. The calculation itself is not the hard part. The data architecture is.
Run that calculation by lead source and the picture changes entirely. A source producing a 43 percent retained revenue rate is not the same asset as a source producing an 81 percent retained revenue rate, regardless of what the cost-per-lead says. The blended marketing cost target only makes sense when the denominator is retained revenue, not signed contract value.
Run it by rep and the picture changes again. The sales manager making decisions on close rate alone is managing the front half of the story. Retained revenue by rep tells the complete story, including what happened after the rep left the driveway.
What PE Acquirers Actually Ask For
Private equity activity in the home improvement space has increased substantially. Bath remodeling, roofing, windows, and solar have all seen meaningful platform acquisitions and roll-up activity over the past several years. The operators who have been through a diligence process have encountered a version of the same experience.
The acquirer asks for retained revenue rate by cohort. By source. By rep. By product category. By market. They want to see the retained revenue rate trend over six to eight quarters. They want to understand which lead sources are producing durable revenue and which are producing signed contracts that cancel before the deposit clears.
Most operators cannot produce this data. The CRM shows activity. The accounting system shows cash. The financing platform shows approvals and denials. The operations system shows installs and cancellations. No single system connects all four, and no one has ever built the data infrastructure to produce cohort-level retained revenue reporting.
The diligence process then becomes an excavation. The acquirer's team reconstructs the retained revenue picture from four disconnected data sources, finds the real retained revenue rate, and adjusts the valuation accordingly. Operators who thought they were selling a business running at 89 percent gross margin discover that their retained revenue rate has been running at 74 percent, and that the gap between signed contract value and retained revenue represents a meaningful adjustment to the multiple.
Retained revenue is not a metric acquirers invented. It is the number the business has always been producing. Most operators just never looked at it directly.
The operators who are acquisition-ready are not necessarily the ones with the highest gross revenue. They are the ones who can produce cohort-level retained revenue data on demand, tracked by source and rep over multiple years, with a retained revenue rate that is stable or improving. That data tells an acquirer something a net revenue trend line cannot: whether the growth is structural or borrowed from future periods that will eventually deliver the cancellations the current report has not yet seen.
What Changes When You Manage to Retained Revenue
Every major management decision in a home improvement operation changes when the target metric shifts from net revenue to retained revenue.
Budget allocation. Lead spend is evaluated against retained revenue per source, not cost-per-lead or close rate. A source producing 200 leads at $90 each with a 38 percent retained revenue rate is a different investment than a source producing 180 leads at $110 each with a 79 percent retained revenue rate. The second source costs more per lead and produces less volume. It produces significantly more retained revenue per dollar spent. Standard reporting does not surface this. Scaling spend into the first source amplifies the leak.
Compensation structure. Reps are paid on retained jobs, not signed contracts. A rep who closes 44 percent and cancels 34 percent is producing 29 percent net retained output. A rep who closes 26 percent and cancels 4 percent is producing 25 percent net retained output. The difference in compensation under a net-revenue comp plan is significant. The difference in actual retained revenue contribution is negligible. The Retention-Adjusted Close Rate is the rep-level version of the same calculation.
Hiring benchmarks. When you are hiring against your current top rep's close rate, you are replicating a metric that may not reflect actual retained revenue contribution. The hiring benchmark changes when the target is retained revenue per demo rather than close rate per demo.
Capacity decisions. Adding capacity to serve higher lead volume only makes sense if the retained revenue rate supports the incremental spend. A business running at a 74 percent retained revenue rate and considering a 20 percent volume increase needs to model the incremental retained revenue against the incremental cost, not the incremental signed contract value.
What Healthy Looks Like
Strong operators frequently retain above 88 cents of every signed dollar by the time a cohort settles. Exclusive lead sources and high-intent inbound programs often run higher. Shared platforms and high-volume aggregator sources typically run lower, sometimes significantly lower, which may still be acceptable depending on what the fully loaded cost per retained job actually is from those sources. The number that matters is not the retained revenue rate in isolation. It is the retained revenue rate against the cost required to produce it.
A blended retained revenue rate running persistently below 75 percent is not a bad month. It is a structural condition. The gap between signed contract value and retained revenue is concentrated somewhere specific: a source, a rep, a product category, a stage in the post-sale process. It does not distribute evenly and it does not self-correct. It compounds quietly while the net revenue report continues to show acceptable numbers.
The operators who have built visibility into this are not necessarily better at marketing or sales. They are better at reading the complete picture. They know which sources hold and which leak. They know which reps produce durable revenue and which produce signatures that do not survive the cooling-off window. They can answer the questions an acquirer will ask before the acquirer asks them.
Net revenue is the score. Retained revenue is the diagnosis of why the score is what it is. Most operators are managing the score. The ones with the clearest picture of where they are going are managing the diagnosis.
See your retained revenue rate by source, by rep, and by period.
Show Me My Revenue Gaps →