← Intelligence Hub

Marketing Intelligence

The Lead Was Cheap. The Revenue Wasn't.

Cost per lead measures where your money went. Revenue per lead source measures whether it worked.

Derwin Lucas, Verisyn HQ · June 2026 · 10 min read

The marketing budget meeting happens the same way in most home improvement operations. Someone pulls the cost-per-lead report. The channels with the lowest CPL get defended. The channels with the highest CPL get questioned. A budget allocation decision gets made. And that decision is based almost entirely on the wrong number.

Cost per lead tells you where your money went. Revenue per lead source tells you whether it worked. The case against optimizing for CPL alone is made in full in Why the Lowest CPL Is Costing Home Improvement Contractors the Most Money. These are not two versions of the same question. They are different questions entirely, and most operators only ever answer the first one.

The measurement stops at the lead. That is exactly where the useful measurement begins. What happens after a lead enters the pipeline, whether it sets, runs, closes, survives rescission, and installs, determines what that lead was actually worth. A $45 Facebook lead that never installs costs more than a $110 Google lead that closes on the first sit at $22,000. CPL cannot tell you which one you have. Only 90-day revenue attribution can.

What the Channel Costs Actually Are

Before the attribution argument can be made fully, the channel benchmarks need to be on the table. These are the CPL ranges that reflect current patterns across bath remodeling, roofing, windows, and siding in competitive markets. They are context, not conclusion.

Google Search Ads

$35–$110

High Intent

Homeowner actively searching a service. Roofing skews to the high end ($150+ in competitive metros). The intent signal is strong. The cost reflects it.

Google Local Services Ads

$28–$55

High Intent

Exclusive, pay-per-lead, Google-verified. Close rates 30–45% reported. The lowest cost-per-closed-job of any paid channel when the sales process is dialed in.

Facebook / Meta Ads

$20–$60

Interrupted Intent

Homeowner was not searching. The ad found them. Lower immediate close rate (8–18%), but 90-day nurture sequences lift that figure significantly. Requires follow-up infrastructure.

Aggregators (Angi, HomeAdvisor)

$85–$150

Shared, Competed

Same lead sold to 3–5 contractors simultaneously. Close rates 6–12%. Blended cost per booked job reported as high as $2,500. The CPL looks manageable. The math downstream does not. See The Hidden Cost of Shared Leads and Shared Lead Platforms Are Not a Lead Source Strategy.

SEO / Organic

$30–$90 effective

High Intent

Slow to build, lowest cost at scale. Effective CPL drops as content compounds. The 12-month investment that most operators defer is the one that would reduce paid dependency most significantly.

Referrals

Near zero direct cost

Highest Intent

Highest close rate of any channel. Cannot be scaled directly. Can be systematized through follow-up, review generation, and referral programs. Most operators leave this channel unmanaged.

These CPL ranges are the numbers most operators use to make channel allocation decisions. They are also the numbers least likely to predict which channel is actually producing revenue. The cost is visible. The return is hidden. That is the measurement problem.

The Number That Actually Answers the Question

The question operators are really asking when they compare channel costs is: which channel should I spend more on? Cost per lead cannot answer that question. It can only tell you which channel charges less per inquiry. Whether that inquiry converts to an appointment, whether the appointment converts to a contract, whether the contract survives to an install, and at what average contract value, are all downstream of the CPL and invisible to it.

The number that answers the actual question is revenue per lead issued by source, calculated over a trailing 90-day window. It is the total installed revenue attributed to a channel divided by the number of leads that channel generated in the same period. It collapses every downstream variable, set rate, run rate, close rate, average contract value, and cancel rate, into a single output figure that makes channel comparison unambiguous.

The CPL Inversion

Facebook  /  "The Cheap Lead"

$45 CPL

Looks like the efficient choice

$180

Revenue per lead issued (90-day) / 4.0x return

Google Search  /  "The Expensive Lead"

$110 CPL

Looks like the costly choice

$520

Revenue per lead issued (90-day) / 4.7x return

The Google lead costs 2.4 times more. It produces 2.9 times the revenue. The return ratio is higher. The absolute revenue per dollar of spend is higher. Every metric that actually measures what marketing is supposed to do, generate installed revenue, favors the channel the CPL comparison said was expensive. The operator who allocated away from Google Search because of the CPL number was optimizing for the wrong variable and paying for it downstream without a report to show it.

This inversion is not hypothetical. It is the pattern that surfaces consistently when operators build full-funnel attribution for the first time. The channel they had been scaling because the CPL was low is frequently the channel with the conversion problems that make cheap leads expensive by the time they install. The channel they had been questioning because the CPL was high is frequently the one producing the strongest revenue per lead in the system.

Why the Attribution Stops at the Wrong Place

Most operators have some version of lead source tracking. The channel that generated the inquiry is recorded when the lead enters the CRM, or when the setter books the appointment, or sometimes not at all. What almost universally does not exist is the connection between that source tag and the final installed revenue number.

The lead comes in tagged as Facebook. It gets set. It runs. It closes. It cancels. Or it installs. At every stage after the initial source tag, the channel attribution is typically invisible. The pipeline report shows revenue by rep, by product, by territory. It does not show revenue by originating channel over a 90-day window. The CRM has the data to produce that report. Almost no operator has built it.

The result is a measurement system that is precise about cost and vague about return. The marketing platform reports CPL with three decimal points. The business owner has no idea what revenue each platform is actually producing. They make channel allocation decisions based on lead volume and gut read, which is a preference masquerading as a strategy.

90 days

The attribution window that matters in home improvement.
A lead generated today may not install for 8 weeks. Shorter windows cut the data before the revenue arrives.

Building the Attribution System

Full-funnel attribution in home improvement does not require sophisticated technology. It requires three things configured correctly and used consistently.

First: source tagging at intake. Every lead that enters the pipeline needs a source field that is populated at the moment of first contact. Not estimated later. Not defaulted to the most recent campaign. Populated at intake from the actual originating channel. A lead that comes in from a Facebook form should be tagged Facebook at the moment the form is submitted, automatically if possible, manually if not. This is the foundation. Without it, nothing downstream is attributable.

Second: the source tag travels with the lead through every stage. Set, run, close, cancel, install. The source field that was populated at intake needs to remain visible and queryable at every subsequent stage of the pipeline. An operator who can see that the lead closed but cannot see which channel generated it has lost the attribution the moment the deal moved past the first stage. The tag needs to be a persistent field, not a note.

Third: a 90-day revenue report by originating source. Once the tagging infrastructure is in place, the report is straightforward: total installed revenue, grouped by originating source, for leads generated in the trailing 90-day window. Divide installed revenue by lead count per source. That is revenue per lead issued by channel. That is the number that answers the allocation question.

Channel	Leads (90 days)	CPL	Total Spend	Installed Revenue	Revenue / Lead	Return Ratio
Google Search	38	$95	$3,610	$161,400	$4,247	44.7x
Google LSA	51	$42	$2,142	$189,300	$3,712	88.4x
Facebook / Meta	112	$34	$3,808	$179,200	$1,600	47.1x
Angi / Aggregator	44	$110	$4,840	$39,600	$900	8.2x
Referrals	19	$0	$0	$133,000	$7,000	Uncapped

This table is hypothetical but calibrated to patterns that emerge consistently when operators build this report for the first time. The insight that almost always surprises: the aggregator channel, which has the highest CPL and the widest volume, is producing the lowest revenue per lead by a factor of four or more. The operator running this channel because it generates lead volume is spending $4,840 to produce $39,600 in installed revenue. The Google LSA channel, which generates a third of the lead volume at $2,142 in total spend, is producing $189,300. The allocation decision that math makes is not subtle.

The referral channel, which has no direct cost, produces the highest revenue per lead in the system. Most operators treat it as a passive outcome rather than a managed channel. A systematic referral program, a post-install follow-up sequence designed to generate reviews and referrals, a structured ask to satisfied customers, does not require ad spend. It requires operational discipline. The return on that discipline, measured against any paid channel, is not comparable.

What the Attribution Report Reveals Beyond Channel Mix

Revenue per lead by source is the primary output of full-funnel attribution. But the same infrastructure that produces it surfaces two additional readings that matter.

Channel-specific funnel behavior

When source attribution travels through every stage, an operator can see not just the final revenue per channel but where each channel's leads are converting and where they are dropping. A Facebook lead that sets at 38% but runs at 65% is telling you something specific: the interest was real enough to book, but the appointment quality was not strong enough to hold. That is a setter script problem for that channel's lead profile, not a channel problem. The funnel gate benchmarks show what healthy set and run rates look like at each stage. The fix is not to abandon Facebook. It is to adjust the qualification conversation for a lower-intent lead type.

A Google search lead that sets at 58% and runs at 84% but closes at 19% is telling you something different: the prospect arrived, showed up, sat through the demo, and didn't buy. That is a presentation problem, or a pricing misalignment for the type of project that channel tends to attract. The yield-per-appointment calculation by channel surfaces this pattern clearly. Again, the channel is not the problem. The system receiving the channel's leads is.

The lag the 30-day report misses

Home improvement is not a same-day close category for most project types. A lead generated from a Facebook ad today may not install for 10 to 12 weeks: the set takes a few days, the appointment runs the following week, the contract is signed, the rescission window passes, materials are ordered, the installation is scheduled. The revenue from that lead does not appear until week ten.

An operator evaluating Facebook performance on a 30-day window is looking at a report that systematically excludes most of the revenue that channel generates. They see high lead volume, high spend, and low visible revenue. They cut the budget. The channel was working. The measurement window was wrong. Why Month-to-Month Lead Data Is Lying to You covers this exact pattern. A 90-day attribution window is the minimum that captures the full revenue cycle for mid-ticket home improvement. For high-ticket projects with longer sales cycles, 120 days is more accurate.

Never evaluate a home improvement marketing channel on a window shorter than the average time from lead to installed revenue in that category. Evaluating a remodeling channel on 30 days is like reading the first three chapters of a book and concluding you know the ending. The story is still being written in the pipeline.

Build the 90-day revenue-per-lead-source report. Run it monthly. The channel that looks expensive at CPL and rich at revenue per lead is the channel that deserves more budget. The channel that looks cheap at CPL and thin at revenue per lead is the channel that is consuming spend and producing noise.

The Allocation Decision the Math Makes

Once the 90-day revenue-per-lead-source report exists, the marketing budget allocation question answers itself. Not perfectly, because lead volume per channel is finite and scaling spend does not always scale results proportionally. But the direction of the allocation becomes a data reading rather than a preference.

The operator who knows that Google LSA produces $3,712 in revenue per lead at $42 CPL, and that Facebook produces $1,600 in revenue per lead at $34 CPL, has a clear relative value statement. Google LSA produces 2.3 times the revenue per lead at 1.24 times the cost. The math favors LSA until the volume ceiling on that channel is reached. At that point, the next dollar of budget belongs on Google Search, not on the aggregator channel producing $900 per lead at $110 CPL.

This is not complex analysis. It is arithmetic applied to data most operators already have in their systems but have never connected into a single report. The technology requirement is a CRM with a persistent source field and a report that groups installed revenue by that field over a 90-day window. That is a configuration task, not a software purchase. How to Calculate Your True Cost-Per-Acquisition by Lead Source walks through the mechanics of building this report from CRM data.

How to Read a Lead Source the Way an Investor Reads a Portfolio and Cheap Leads. Expensive Revenue. develop the full investment-lens framework. The operators who are making accurate channel allocation decisions are not spending more on marketing. They are spending the same budget in a different distribution, guided by a report that took a few hours to configure and is now running automatically every month. The operators who are not running this report are making the same allocation decisions by feel, with the same budget, and wondering why the return is inconsistent.

Build This Report

Open your CRM. Find every lead generated in the last 90 days. Group them by originating source. For each group, sum the installed revenue from contracts that have cleared to installation in that same window. Divide by lead count. That number, revenue per lead by source, is the only channel comparison that accounts for everything that happens after the lead is generated.

If your CRM does not have a persistent source field that travels with the lead through every pipeline stage, that configuration gap is the first thing to fix. The report cannot exist without the data. And without the report, your channel allocation is a preference, not a strategy.

Revenue Intelligence

The Revenue Visibility Stack assessment surfaces which constraint type is active in your operation. Eight questions. Specific output.

The Verisyn HQ Intelligence Hub applies these frameworks to your operation. Diagnostic tools, benchmark maps, and operator-level analysis.

Browse the Intelligence Hub →

Continue Reading

Browse the Intelligence Hub →