The eight checks
Each one is the same three questions: what to look at, what good actually looks like, and what it costs you when it's missing. Most roofers pass the first few and fail the last two — which is the opposite of the order that matters.
Where does the click actually land?
Click your own ad from your phone. Write down the URL it lands on. Then read the ad's headline and the first line of that page back to back, out loud.
The click lands on a page built for that one ad, and the page repeats the promise the ad made — same job, same city, same words. If the ad said storm damage inspection, the page says storm damage inspection.
A homepage has to serve everyone: the roof replacement you want, the gutter job you don't, and the About page. A homeowner who clicked for one specific thing has to go find it again — and most won't. You paid full price for that click either way.
Where is the form, and how long is it?
Land on the page on your phone and don't scroll. Is there something to act on in that first screen? Then count the form fields. For each one, ask who in your office actually uses that answer.
Something actionable above the fold, and the shortest form that still lets you qualify. Every field earns its place. Anything you could ask on the phone in ten seconds shouldn't be a barrier before you have the lead at all.
Every extra field is another chance to quit, and a form below the fold is a form most visitors never see. You're not losing people who didn't want a roof. You're losing the ones who did and got tired.
Can a homeowner get a number without waiting for you?
Pretend you're the homeowner. Can you get any kind of price — a range, a ballpark for a roof like yours — without submitting a form and waiting for a callback?
A real, defensible range tied to their actual roof, produced on the spot. Contact details unlock the detail and the appointment; they aren't a toll booth on the entire answer.
If your page says "call for a free estimate" and the next result gives a number in thirty seconds, you're asking for more and offering less. Search results have started sorting on this too — see what homeowners expect before they call you.
Checks 4–8
Get the full checklist
The remaining five checks are the ones almost everybody fails.
One email. No sales calls. Unsubscribe in one click.
The driveway test: phone and speed
Open the page on your own phone, on cell service, not office wifi. Count out loud until it's usable. Then try to tap the phone number and fill the form one-handed, the way someone standing outside would.
Usable in about the time it takes to say "one thousand one, one thousand two." Buttons a thumb can hit. No pinch-zoom, no sideways scroll, and no popup covering the form you want them to fill in.
A homeowner standing in the driveway looking at a missing shingle is the best lead you'll get all day. A slow page loses them before it finishes loading — and you're billed for the click regardless.
Is anything tracking the calls?
Is the number on your ad landing page different from the number on your truck? For any call you took last week, can you say which ad or which page produced it?
A tracking number on the ad-driven page (your real number everywhere else), calls logged with their source, and somebody actually looking at that log — not just collecting it.
Roofing is a phone business. If calls are invisible, the ad that drives calls and few form fills looks like your worst ad. Sooner or later you turn off the thing that was working.
Does a conversion actually fire?
Submit your own form. Load the landing page and the thank-you page with Google Tag Assistant or Meta's Pixel Helper open. Does a conversion event fire on the submit — or only a pageview?
The tag is on both pages, a conversion fires on the submit, and it's the same event your campaign is set to optimize toward. A tag firing into an event nobody bids on is decoration.
With nothing firing, the platform optimizes for the cheapest clicks it can find. That isn't a bug — you asked for clicks, and it is extremely good at finding cheap ones.
Does anything report outcomes back to Google or Meta?
A lead becomes an inspection. An inspection becomes a signed contract. Ask the plain version: if you signed a job on Tuesday, where does that fact go? If the honest answer is "into the CRM and nowhere else," this one is a fail.
Booked and sold outcomes flow back to the ad platform with the job value attached, so bidding moves toward homeowners who look like people who actually bought. Google's version is Enhanced Conversions for Leads — the mechanics are in offline conversion tracking, in plain English.
This is the one that quietly caps all the others. Without it, a junk lead and a signed roof are the same single "conversion," so the platform keeps buying more of whatever was cheapest. Fix checks 1–6 and you get better leads. Fix this one and the ads get better on their own, every month, without you.
How long until a lead hears from a human?
Have someone you know fill in your form at a genuinely awkward hour — Saturday afternoon, Tuesday evening. Time it. How long until a text, a call, anything?
An automatic text or call inside the first minute or two, before anyone in your office has even seen the lead — then a human follow-up the same day. They should hear from you while your page is still open on their phone.
A homeowner filling in your form has usually filled in two or three others. Whoever answers first gets to frame the job, the price and the timeline. Every hour you wait, you're paying to hand a lead you bought to a competitor who didn't.
Now count your passes
What to do with your count
A failing count isn't a verdict on your ads. It's a map of where the money leaks between the click and the contract. Take the three cheapest fixes, do them, then re-run this list in a month — same phone, same cell service, same awkward hour.
And before you spend anything fixing it, find out what the leak is worth: put your own spend, lead count and close rate into the cost-per-job calculator and see what a signed job costs you today.