How Do You Know Which of Your Leads Came From Marketing?
Short answer: you know because you set it up to be knowable before the leads came in. There is no report I can run in August that reconstructs where your July leads came from. If the tracking wasn't in place, that month is gone, and no amount of digging around in the dashboards afterward will bring it back.
Setting it up isn't complicated. Most of it you can do yourself in an afternoon, and the piece that usually pays for itself fastest takes about fifteen minutes inside your Google Ads account. Almost nobody does it, though, and I can tell you exactly what that costs, because a client came to me spending $25,000 a month on Google Ads with a dashboard that looked fantastic.
The $25,000 Story
His conversions were coming in around $15 apiece, which is the kind of number that makes an owner want to double his budget.
Then I opened the conversions table. Most of what Google was counting as a conversion turned out to be people tapping "get directions." Not phone calls, not form fills, not booked jobs. Map taps.
We fixed what was being counted, cut spend hard, and rebuilt the campaigns around the actions that actually come before money changes hands. He now spends about $3,000 a month on Google and makes more from it than he did at $25,000. Had nobody ever opened that table, he'd have spent a few hundred thousand dollars over a couple of years buying taps on a map.
That's the most expensive version of this problem I've run into, but the underlying condition is everywhere.
What Most Accounts Look Like Under the Hood
Whatever came standard, and nothing past it. The default conversion tracking Google Ads switched on by itself. A Meta pixel somebody pasted in three years ago. Maybe a Google tag on the site.
I have yet to take over an account that was tracking individual forms, individual pages, individual phone numbers, and individual campaigns.
That distinction is the entire game. Knowing that some of your leads came from Google is not tracking, it's a guess with a logo on it. Knowing that the form on your fence installation page brought in eleven leads last month, that seven of them closed, and that they all came from one campaign is tracking. One of those tells you what to do on Monday. The other just makes you feel like you're paying attention.
Your Phone Is Where This Falls Apart
Home service is still a phone-first business. Most people who need a plumber or a roofer are going to call you, and I'd add one wrinkle to that: when you give people the option to actually book a time online instead of submitting a vague interest form, a good number of them will take it. But the phone is still where most of your revenue walks in, and a phone call is the hardest thing in your business to trace.
Fixing it costs less than a single service call. Dynamic number insertion, or a pool of numbers, swaps the phone number displayed on your website depending on where that visitor came from. CallRail, GoHighLevel, and OpenPhone all do it. One tracked number on your Google Business Profile, another on your Google Ads landing page, another on your Meta landing page, and suddenly every ring tells you something it wasn't telling you before.
I turned this on for a client and found that more than half of their phone calls were coming from the Google Business Profile alone. Before that, every one of those calls had been landing in a mental bucket labeled "people just call us." That one number changed how we spent their whole budget.
Follow the Lead All the Way to the Money
A lead is not a result. Push lead source into your CRM, then tie it back to closed revenue in Jobber or ServiceTitan, and look at what the jobs were actually worth.
Hardly anyone runs that report, and it's the one that changes decisions, because the channel producing the most leads is frequently not the channel producing the most revenue. Cheap leads that never close will put you out of business slower and more pleasantly than expensive ones that do, but they'll still put you out of business.
What a Lead Is Actually Worth to You
Most owners judge cost per lead against a number they invented. A hundred bucks a lead sounds like too much. Too much compared to what?
Here is the math that settles it.
What a lead really costs you, after accounting for the ones you don't win:
Customer acquisition cost = cost per lead ÷ close rate
What a job is really worth to you:
Gross profit per job = average job value × gross margin
If your work is recurring, or you can upsell into a membership, use average lifetime value in place of average job value. For small-ticket trades that swap matters more than anything else on this page.
Your ceiling:
Max acquisition cost = gross profit per job × the share of that profit you're willing to spend to win the customer
Max cost per lead = max acquisition cost × close rate
And the health check:
Acquisition cost ÷ gross profit per job = the share of each job's profit going to marketing
Run real numbers through it. A $100 lead at a 50% close rate is a $200 acquisition cost. A $5,000 job at 30% margin leaves you $1,500 of gross profit. Marketing took 13% of it, you kept $1,300, and you should buy every one of those leads you can get your hands on.
Now put a land prep company through the same equation. Thirty thousand dollar jobs at 30% margin, so $9,000 of gross profit. Even at a 25% close rate, that same $100 lead costs $400 to convert, which is under 5% of the profit on the job.
I ran into exactly that with a land prep client. We were underfunding Google Local Services because the cost per lead looked high next to what other trades pay. It wasn't high at all, and we were leaving real money on the table by not spending more. When your average job runs into the tens of thousands, you should be thrilled to pay $100 for a solid lead. Big-ticket contractors ought to get nervous when their cost per lead looks cheap.
Flip the numbers and you get the opposite answer. A $500 pool tile cleaning job at 50% margin is $250 of gross profit. That same $100 lead at a 40% close rate costs $250 to convert, so you broke even and paid yourself nothing. Put that customer on quarterly service for two years, though, and you're now working against a few thousand dollars of lifetime revenue rather than $500, and the identical lead becomes one of the best buys on your board.
Same lead, same price, opposite decision, which is why the equation beats anybody's opinion about what a lead should cost.
As for what share of gross profit to cap acquisition at, there's no universal number and I don't set one. It depends on how badly you need volume and what you can stomach. A company staring at an empty February should be willing to pay considerably more than one that's booked into next quarter.
Some of It You Will Never Know
Most agencies won't say this part out loud.
Somebody sees your Facebook ad in March. Drives past your truck in April. Asks a neighbor in May. Googles your business by name in June and calls. Every platform in that chain will cheerfully take full credit for the job.
A percentage of your leads will always be unattributable, and that percentage moves depending on how many channels you're running. You can narrow the gap quite a bit by giving each channel its own phone number and its own landing page with its own form, which is how we separate Google from Meta. You will not get to a hundred percent, and anybody who promises you otherwise is selling something.
When a client pushes back on the numbers, we look at total return on ad spend. If the whole machine is producing and they're making good money, the argument about which channel deserves the credit tends to get quieter. And when we really do need to know, we test.
Testing a Channel Without Wrecking the Data
You've seen a brown ring on a ceiling under a second-floor bathroom. Could be the supply line, could be the wax ring, could be the shower pan, could be a nail somebody put through a pipe three years ago. Nobody good tears out the whole ceiling to find out.
You cut the smallest hole that tells you the most and rule things out one at a time.
Say you're spending $4,000 a month on Google and $2,000 on Meta, and you suspect
Meta is where the money is coming from. Bump Meta by 20%, so $400, and leave everything else exactly where it sits. Give it two to four weeks. Watch leads, closed jobs, and revenue.
If revenue moves, you found something. If it doesn't, Meta is probably doing awareness work rather than driving demand you can bank, so bring the spend back down and run the same test on Google with a 20% bump, which in this case is $800.
One channel at a time, and never both at once. Raising one while lowering the other is worse still, because now you've changed two variables and the result can't tell you anything. That's tearing out the ceiling.
Five Things You Can Do This Week
Do them in this order.
Clean up your conversion actions in Google Ads. Open the conversions table and look at what's marked primary. Demote directions requests, page views, outbound clicks, and anything the platform created on its own. Primary should be calls over a minimum duration, form submissions, and booked appointments, and nothing else. Fifteen minutes, and it's the fastest money on this list.
Put a tracking number on your Google Business Profile. If you only do one thing here, do this. Half your calls may already be coming from there and getting filed under "organic."
Add a required lead source field in your CRM, and make it a locked dropdown rather than free text, or you'll wind up with forty spellings of Facebook and a field nobody can report on. Then have whoever answers the phone ask people how they found you and actually write it down.
Put UTMs on every link you personally control. The website link on your Google Business Profile, your email signature, your social bios, the QR code on the truck.
Once a month, pull closed jobs and revenue by lead source. Revenue, not lead count.
Where Doing It Yourself Runs Out
That list gets you most of the way, and I'd much rather you do it yourself than not do it at all.
It gets harder when you start pushing closed revenue back into the ad platforms so the algorithms optimize toward booked jobs instead of form fills. Sizing number pools correctly so dynamic number insertion doesn't break and start showing the wrong number to the wrong visitor takes some care. And running a clean isolated test when leads dip has less to do with technical skill than with having somebody outside your business who isn't panicking alongside you.
You would never bid a job without knowing your material cost, your labor hours, and what you billed. You'd let a foreman go if he couldn't tell you whether a job made money. Marketing deserves that same standard, and in most home service businesses it isn't getting it.
Not sure what's actually producing in your business? Book a free 30-minute strategy call. We'll look at what you're spending, what you're tracking, and where the leaks are, and you'll walk away with a clearer picture of your numbers whether or not we end up working together.
Comments