How to tell which ads actually bring you booked jobs

We hear a version of this story every week. An owner is paying 3 places at once. Google, Yelp, and the Facebook guy. Add it up and it is about $900 a month.

The phone rings. The trucks stay busy. So something in that stack is working, and that is exactly the problem, because nothing in the stack says which. The way owners put it to us usually lands close to this: "Somebody's ads are bringing the phone calls, I just couldn't tell you whose, so I keep paying all of them."

That is not carelessness. It is a measurement problem wearing a costume. Cancel the wrong line and the phone goes quiet in a week. Cancel nothing and you keep feeding the dead ones forever. So the owner keeps feeding them, because being wrong costs more than being unsure. Uncertainty costs the same every month, quietly, and never sends a bill that says what it is for.

One tracked line per channel. That is the whole trick.

Every place you spend a marketing dollar gets its own phone number, and each one forwards straight to your shop line. Your customer dials, your phone rings, nobody notices a thing. Your real number stays on your Google listing, following Google's own documented call-tracking pattern, and the whole setup reverses in 15 minutes.

Then the counting starts. Every booked job gets matched back to the number that brought the call. On Monday you get one email: qualified calls per channel, confirmed booked jobs per channel, and cost per booked job for each. Not clicks. Booked jobs, sitting next to what you paid for them. No dashboard, one email you can read in the truck.

The scoreboard comes with rules built in, because a scoreboard that flatters itself is worse than none. Booked-job matches are always labeled as a floor, never inflated: the phone that calls is often the spouse's, not the one on the invoice, and we would rather undercount than take credit we did not earn. When the report misses one, every call row carries a one-tap "this call became a job" button. No channel is called dead without 10 or more qualified calls, zero matches, and your own spot-check against the recordings linked in the email. A repeat-and-referral bucket keeps the report adding up to your actual week.

Somewhere a rack of numbers counts rings all night while the trucks sleep in the yard. On Monday it hands you a sheet of paper. That is the entire romance of it.

Tracked lines go live in 1 to 2 days. The first Monday email arrives inside week one. At 60 days every channel has a verdict, with cost per booked job side by side, and the dead dollars move onto the channel that books, so the same budget buys a better phone. An optional signs module tracks 50 yard signs by zone, with a monthly move-these-signs list.

Setup is $997. The care plan is $299 a month. The yard-sign zone module adds $497 setup and $99 a month for up to 10 tracked zone numbers. That is the whole price, and it is printed on the page.

Here is the break-even, not the best case. Year one is $997 plus 12 months at $299, which is $3,588, for $4,585 total. At $225 gross per booked job, $4,585 divided by $225 is just over 20 booked jobs, so call it 21. That is 21 extra booked jobs spread across 12 months, fewer than 2 a month, and the year has paid for itself.

Now an example. Say Yelp is taking $350 a month and shows zero booked jobs in 60 days, while Google books 9 at about $33 each. Cutting the dead channel is the demo. The payday is the move. Take that $350, put it where the booking happens, and count conservatively: 3 extra booked jobs at $225 gross is $675 a month against $299. That clears the $997 setup inside 3 months.

Not one extra dollar of budget. Same money, standing somewhere else.

Under $600 a month in ad spend, this does not pencil. Below that line the report cannot move enough dollars to pay for itself, and we say so before you sign, not after.

Do not buy it if you want us anywhere near your leads. This is measurement only. Calls forward straight through, leads hand off untouched to however you handle them today, and the answering stays yours.

Do not buy it if you want a verdict by Friday. Verdicts land at 60 days, and the dead-channel rule exists precisely so the report cannot talk you into killing an ad that works.

Do not buy it if you need attribution to be exact. It is a floor, on purpose, and it says so on every line.

And do not buy it hoping we will take credit for your signs. A $12,000 roof traced to a sign was coming anyway, since the sign was already in the ground. The signs column buys you a decision instead: inside one season you know which streets make the phone ring, next season every sign stands on a street like those, and the dead ones stop getting reprinted.

You are already paying for this information. About $900 a month, and it does not arrive. 60 days after the lines go live, every channel has a number next to it, and guessing stops being a line item in your budget. The price is printed above. So is the qualifier.