Most roofing companies can count leads but can't say which marketing produced sold jobs. Here's how to find out whether your marketing is working.
A roofing owner sits down with last month's marketing invoices. Google Ads. Local Services Ads. An SEO retainer. The social media guy. A lead marketplace. The sponsorship banner at the high school ball field.
Somewhere north of $15,000.
He knows roughly how many leads came in. He knows he sold eleven jobs.
What he can't tell you is which of those eleven jobs came from which marketing invoice.
So he renews all of it. Again.
That's a revenue leak, and it's an unusual one. Nothing broke.
What he's missing is marketing attribution — knowing which source produced which job. Without it, he can't cancel anything, because any line he cuts might be the one that was working. Nothing comes off the list, and next quarter somebody pitches a new channel, and that goes on the list too.
Here's how that reaches your margin. Every job you sell carries a share of what you spent to get it. Spend $15,000 and sell eleven jobs, and each one is carrying about $1,400 in marketing before you've bought a single bundle of shingles. Let the budget drift up to $20,000 while you're still selling eleven, and that's $1,800 a job.
Nothing about the work changed. The job just got more expensive to win, and that difference eats into your margin.
This article covers why marketing attribution breaks down in a roofing business, what the gap costs you, and how to check your last twenty sold jobs to see whether you have this leak.
Table of Contents
- Key takeaways
- What is marketing attribution for a roofing company?
- Why can't most roofing companies tell which marketing produces jobs?
- Why counting leads leads to bad marketing decisions
- Why "I found you on Google" doesn't tell you anything
- What about the sources no system can see?
- What happens when your source data disagrees with itself?
- What is bad marketing attribution costing your margin?
- How to check your marketing attribution yourself
- What does good marketing attribution look like?
- How does marketing attribution fit into the bigger revenue picture?
- Find your roofing revenue leaks
- Marketing attribution FAQs
Key takeaways
- Most roofing companies can count leads by source but cannot connect sold jobs and revenue back to the marketing that produced them.
- Lead source information usually stops at the lead. The job gets sold and produced in a different system, and nobody ever connects the two.
- Without marketing attribution you can't cancel anything, so the budget only grows — and every job you sell carries more acquisition cost than it did last year.
- "I found you on Google" covers paid search, Local Services Ads, Google Business Profile and organic results, which are four different budget decisions.
- You can check your own marketing attribution with a list of your last 20 sold jobs and a pen. You do not need software to find out whether you have this problem.
What is marketing attribution for a roofing company?
Marketing attribution is knowing where an opportunity came from and keeping that information attached to it all the way through to a sold job. It is not the same as knowing how many leads a channel generated.
The first half is easy, and most roofing companies have some version of it. A call came from the ad number. A form said Google. A homeowner told the office they were referred by a neighbor.
The second half is where it usually falls apart. The source needs to survive the trip through the appointment, the estimate, the sale and the invoice — because revenue is the only thing that tells you whether the marketing worked.
Why can't most roofing companies tell which marketing produces jobs?
Because the source information dies before the money shows up. The lead arrives with a perfectly good source attached. Then the job gets sold and produced somewhere else, and nobody ever connects the two records.
Think about where the information actually lives in a typical roofing company.
The lead comes in through the website, a call tracking platform, an ad account or a lead marketplace. The source exists there.
The job gets estimated, sold, scheduled, built and invoiced in your production or accounting system. The revenue exists there.
Those two systems often have no idea the other one exists. Or they are connected, but the source field was never part of what moves between them.
So at the end of the month you can produce a report on leads by source, and a report on jobs by revenue, and there is no honest way to lay one on top of the other.
This is why owners who suspect they have an attribution problem often try to fix it by building better reports. The reports were never the problem. The information stopped upstream, and no amount of reporting recovers data that was never carried forward.
Why counting leads leads to bad marketing decisions
Lead volume and revenue are different measurements, and the channel that wins on one often loses on the other. When leads are the only thing you can count, leads become the thing you optimize for.
Here is an illustration. These are made-up numbers used to show the shape of the problem, not industry benchmarks.
Say paid search produced 40 leads last month and organic search produced 25. Paid search looks like the better channel, producing well over half again as many opportunities.
Now suppose those 40 paid leads produced 2 sold jobs, and the 25 organic leads produced 8.
At a $12,000 average job, that is $24,000 from paid search and $96,000 from organic. Same month, opposite conclusion.
Without the second half of that picture, the owner does the reasonable thing and puts more money into the channel that generated more leads.
It is worth noticing who benefits from lead counting. The businesses selling you leads report on leads. That is not dishonest — it is what they can see. But it means the scoreboard you are handed measures their product rather than your business.
Why "I found you on Google" doesn't tell you anything
"Google" is not a marketing channel. It is at least four of them, and they have completely different costs.
When a homeowner says they found you on Google, they could mean:
- A paid search ad you are paying for by the click
- A Local Services Ad you are paying for by the lead
- Your Google Business Profile, which costs you nothing but attention
- An organic search result, which is what your SEO retainer is supposed to produce
Those are four different budget decisions. The homeowner has no idea which one they used, and there is no reason they should. They typed in "roof repair near me" and called the first company that looked credible.
This is why self-reported source information alone is not enough. It is useful, and for some sources it is the only thing you will ever have. But when your records are full of contacts whose source is "Google" or "online" or "internet," you have a category, not an answer.
The homeowner's experience matters here too. If your intake is asking a homeowner with water coming through the ceiling to remember exactly how they found you, that question is doing more for your marketing file than for the conversation. There are better places to capture it, and better moments.
What about the sources no system can see?
Some of your best roofing leads leave no digital trail at all, and the only way they get recorded is if a person records them.
A neighbor recommendation. A yard sign two streets over. Your truck parked in a driveway for three days. A knock on the door after a storm. A repeat customer calling the owner's cell phone directly.
No tracking code will ever capture those. And in a lot of roofing companies, those are the sources producing the best jobs at the lowest cost.
I wrote in the article on roofing websites that "How did you hear about us?" is administrative work you are pushing onto the homeowner. That is still true when you are asking it to do a job your systems should already be doing.
The distinction is this: your systems should capture what systems can see, and people should capture what only people can see. Nobody needs to ask a homeowner which ad they clicked. Somebody does need to write down that Mrs. Kowalski called because her sister-in-law had you out last fall.
Which raises a question most roofing owners have never actually answered: where does that information go? If a crew lead hears at the job site that the neighbor wants an estimate, is there a place for that to land? Or does it live in a text message until everyone forgets?
What happens when your source data disagrees with itself?
When a company does have source tracking, the common failure is not missing data. It is several fields telling different stories about the same lead.
I worked on this problem for a business in another industry that had real tracking in place, properly installed, with more source information than most companies ever collect. That was the problem. The system recorded a traffic source, campaign tags from the ad platform, a referral field the customer filled out and a record of how the contact was created.
They did not always agree. A contact could be labeled direct traffic by one field while the campaign tags on the same record clearly identified a paid search click. Both were true in a narrow technical sense. Only one was the answer.
What fixed it was not better tracking. It was a decision: one standardized source for every contact, and a written priority order for which evidence wins when the fields conflict. The strongest available signal is used, the same way, every time. Referral information beats an ad tag. An ad tag beats a generic traffic label. A generic traffic label beats nothing.
Two things make that work, and neither is technical. Somebody has to decide the order, and somebody has to own it when the business changes. Add a channel, change a form, rename a campaign, and the rules need to be checked. A source model nobody maintains quietly goes wrong, and it goes wrong without telling you.
What is bad marketing attribution costing your margin?
It doesn't take money out of a job directly. It makes every job more expensive to win.
Go back to the owner at the start of this article. $15,000 a month, eleven jobs, about $1,400 of marketing cost riding on each one.
Now ask what happens when he can't tell which of those invoices produced work. He can't cut anything, because cutting the wrong line is worse than overspending. So the list only grows. Next spring it's $18,000, then $20,000, and the job count hasn't moved.
At $20,000 and eleven jobs, that's $1,800 riding on each one. Same crews, same materials, same roofs. Four hundred dollars a job gone, before anybody picks up a hammer.
Here's a second way to size it, using your own numbers. Take your total monthly marketing spend, then write down how much of it you could connect to a sold job if someone asked you today.
If you spend $15,000 a month and can confidently trace $5,000 of it to actual revenue, the other $10,000 isn't necessarily wasted. Some of it is certainly working.
The honest statement is that $10,000 a month is being renewed on faith, and that over a year that's $120,000 in decisions you have no way to check.
That's the number worth finding. Not because the money is gone, but because you can't improve a decision you can't see.
How to check your marketing attribution yourself
You can find out whether you have this problem in about 10 minutes, with a list of recent jobs and a pen. Don't start in an analytics dashboard. Start with jobs you actually sold, because that is the part that matters.
Test 1 is the ten-minute one. Work through the rest as you have time.
Test 1: Your last 20 sold jobs
Pull up your last 20 sold jobs. For each one, write four things:
- Customer
- Job value
- Where the lead came from
- How you know that
That fourth column is the test. "It's in the system" is an answer. "Pretty sure it was the Facebook thing" is not.
Count how many of the 20 you can answer confidently. If it is fewer than 18, you have this leak. If it is fewer than 10, the leak is the main thing standing between you and knowing whether your marketing works.
Test 2: Match your spending to your revenue
List every marketing expense from the last 90 days. Every vendor, every platform, every sponsorship, every membership.
Next to each one, write the number of sold jobs and the revenue you can trace back to it.
The blanks are the leak. Not the small numbers — the blanks.
Test 3: Follow one job backward
Pick one sold job from the last month. Start at the invoice and work backward.
- Who sold it?
- Where did the estimate come from?
- Who set the appointment?
- How did that person first reach your company?
- What told you that, and is it written down anywhere?
Note where in that chain you had to ask somebody instead of looking something up. That point is where your source information stops.
Test 4: Check how many leads say "Google" or "online"
Look through your recent leads for source values that are really just categories. "Google." "Online." "Internet." "Website." "Referral," with no indication of who referred them.
Each one of those is an opportunity you paid for and cannot assign to a decision.
Test 5: Ask where offline leads get recorded
Ask your office manager and one crew lead the same question: when somebody says a neighbor sent them, or they saw the yard sign, where does that get written down?
If you get two different answers, or a pause, you have found the gap. Referrals and repeat customers are usually the cheapest revenue in a roofing company and the least likely to be recorded.
Found an attribution gap?
The free Roofing Revenue Leak Check looks for the other places where marketing dollars, leads and revenue may be slipping through your business.
Check your roofing revenue leaks →
What does good marketing attribution look like?
Good marketing attribution means every opportunity has one source, assigned the same way every time, and that source is still attached when the job is sold. It does not require expensive software, and it does not require tracking everything.
Three things have to be true.
One list of sources everyone uses. Not a free-text box. A defined set that reflects how you actually spend money. For most roofing companies, that looks something like:
- Referral from a person
- Repeat customer
- Local Services Ads
- Paid search
- Paid social
- Google Business Profile
- Organic search
- Organic social
- Lead marketplace
- Referring website
- Canvassing or door knocking
- Signs, trucks and offline advertising
- Direct or unknown
Local Services Ads and Google Business Profile get their own lines because you fund and manage them separately. Lead marketplaces get their own line because their economics are nothing like the rest. Referral and repeat customers get their own lines because, in most roofing companies, they're your best revenue.
A decided order for when signals conflict. When a lead has more than one source signal — and good leads often do — you need a rule for which one wins, not a judgment call made differently by whoever is entering it. Write the order down. Follow it every time.
The source survives to the sold job. This matters most, and most companies skip it. If the source is recorded at the lead and lost at the sale, you are back to counting leads.
And one practical note: a source model is not a thing you set up once. Every time you add a channel, change a form, or rename a campaign, you need to check the rules. If you change how source information is collected, check the rules. If you change the rules, test the result.
How does marketing attribution fit into the bigger revenue picture?
Marketing attribution breaks at the beginning of the process and hurts you at the end of it. That gap is why it goes unnoticed for so long.
Lead source → Lead capture → Response → Qualification → Appointment → Estimate → Follow-up → Sale → Measurement
The break happens at lead capture, when the source is never recorded or never carried forward. Nothing looks wrong at the time. Leads come in, appointments get set, jobs get sold, crews go out.
You feel it at measurement, weeks or months later, when you sit down to decide what marketing to renew and find you cannot answer the question.
That distance is why building better reports never fixes this. By the time you are looking at a report, the information you needed was left behind eight stages ago. Reporting can only show you what somebody captured.
The phone is one version of this. In the article on call routing problems that lose roofing jobs, I wrote about tracked numbers that connect the call while losing the record of where it came from. Same leak, one channel. Every channel has its own version.
This is one of fifteen revenue leaks that waste good roofing leads, and the rest of them are worth walking through the same way.
Find your roofing revenue leaks
You may not need more leads. You may need to find out what's happening to the opportunities and the marketing dollars you already have.
The free Roofing Revenue Leak Check looks across your lead-to-revenue process to find where opportunities and information may be getting lost — from how homeowners find and contact you through response, appointments, estimates, follow-up, sales and measurement.
The goal is straightforward: find out whether you're getting everything you can from the opportunities you're already generating.
Check your roofing revenue leaks →
Marketing attribution FAQs
What is marketing attribution for a roofing company?
Marketing attribution means knowing where each roofing opportunity came from and keeping that information attached through the appointment, the estimate and the sold job. Counting leads by channel is only half the story. Attribution is complete when you can tie revenue back to the marketing that produced it.
How do I track where my roofing leads come from?
Use tracking to capture what systems can see, such as calls, forms, ads and website visits, and give your office and field teams a defined place to record what systems cannot see, such as referrals, repeat customers, yard signs, and door knocks. Both halves are necessary. Most roofing companies have some of the first and very little of the second.
Should I ask homeowners how they heard about my roofing company?
Ask when it captures something your systems genuinely cannot see, such as a referral, a yard sign or a truck someone noticed. Do not rely on it for digital sources, because homeowners cannot reliably distinguish a paid search ad from a Local Services Ad from an organic result, and there is no reason they should.
Why do so many roofing leads show up as "Google"?
Because Google is several channels at once. Paid search, Local Services Ads, Google Business Profile and organic search all look like Google to a homeowner, but each has a different cost and requires a different decision from you. A source value of "Google" tells you a category, not an answer.
How can I tell if my roofing marketing is actually working?
Connect each marketing expense to sold jobs and revenue rather than to lead counts. Start with your last 20 sold jobs and see how many you can confidently trace back to a source. If you cannot trace most of them, you are measuring activity rather than results, and lead volume will keep pointing you toward whichever channel produces the most of it.
Comments