He was paying for clicks that landed on a dead page.
Epic Glass & Tinting, an auto glass and window tint shop in Klamath Falls, Oregon, a town of about 22,000. The owner was good at the work and known for it: the most reviewed shop of his kind in the county. The work was never the problem.
His host had suspended the website. The domain served an error page, the certificate was untrusted, and the ads he was paying for every month were still pointing at it. His Yelp page sat unclaimed at 4.4 from nineteen reviews while his Google profile carried 4.6 from 157, and the AI assistants were reading the wrong one.
This is the eleven weeks after. Almost everything below is a screenshot of Google, of his own Google accounts, or of his own records, each one dated.
56quote requests through the site
59taps on his phone number
223clicks from Google search
#1web result on 11 of 12 service searches, October 7
Quote requests counted off the actual enquiries in his inbox, August 4 to October 22, with tests and spam taken out. Phone taps from Google Analytics; tap tracking started August 14, so the first month is undercounted. Google clicks from Search Console, August 13 to October 20. There is no “before” for any of this: the month before, his domain served an error page and Google had nothing to show.
July 31, before. “windshield replacement klamath falls”: his pin is on the map, he isn’t on the list.August 22, three weeks in. Same search, same width: he’s on the list.
★★★★★
“Steve built us an awesome website! Our old site was outdated. The new one shows customers that we are professional. He is continually editing the site to make it better. The site was up within like 2 days. We couldn’t be happier with the results!!”
David Pratt · owner, Epic Glass & Tinting · his Google review of my work
How I got the job
He didn’t take the first meeting. Or the second.
Early June
Before I ever walked in, I researched the shop and the market and built a demo of what his site should be.
June 24
First visit. I showed him the security warning his website was showing customers. He was already trying to build a new site himself.
July
Came back twice more: a written plan for his shop, then a sit-down on July 30.
August 4
He signed, month to month. Asked why he took the meeting: “Persistence. I kept showing up.”
Day zero
What was broken when I started.
Late July: epicglass.com. His paid ads were still sending people here.
378 people tapped “Website” on his Google listing between March and August. They landed on a site marked “Not Secure,” and from late July on, the suspension page. His own Google Business Profile data.
Yelp, July 31: unclaimed, 19 reviews, one broken photo. He had 157 on Google.
Before building anything
I read the market, and I read his calendar.
The shops ahead of him on Google’s map had fewer reviews and higher ratings. He already had the most reviews of any glass shop in town, so the lever was turning reviews into calls and keeping new ones coming, not more pages.
Nobody owned one message. One rival sells thirty years local, another sells cheap, and nobody could say tint, glass and the camera calibration in one trip. No other Klamath Falls shop I found does all three in-house. That became his headline. On the windshield search, Safelite’s ad sits above the map; the local shops mostly aren’t buying.
I went through sixteen months of his job calendar, every entry. His goal was “better winters.” His own calendar said February was his busiest month and the dips were August and November, so I didn’t sell him a winter campaign. Three of every four vehicles on it were trucks and SUVs, so the site speaks to work trucks in a mountain town, not sports cars.
The website
Twenty-three pages, live in about two days.
Homepage, desktop.Homepage, phone, with Call and Quote buttons that stay on screen.
23 pages live: 10 service pages, 8 articles answering what customers ask before they call, plus about, privacy and terms.
A quote form that emails him every request, with the customer’s address set to reply.
A page for every service he actually sells, because Google cannot rank a service it has never been told about.
His listings and his reviews
None of it is clever. All of it was missing.
Google Business Profile: main category changed from “Glass repair service” to “Auto glass shop,” the services filled in, the website link fixed.
Yelp: claimed and rebuilt. Saturday hours corrected, a windshield category with its services added, photos from 1 to 6.
Bing: set up. Better Business Bureau: submitted; every glass rival had an A+ and he had no listing at all.
Review replies written in his voice and posted, including the one-star ones.
His domain secured and verified paid through 2031, because a lapsed domain is a business that disappears overnight.
The counter sign: one scan to leave a Google review. Ordered for his counter.
The uncomfortable part
What his ad company was actually delivering.
He was paying a geo-targeted ad company about $550 a month, his figure. I got into their dashboard and measured it against his own Google data.
Their report: 2,704 people reached. Audience: 100% male.
Last 30 days
His ads
Free Google search
Visits to his website, by his site’s own count
95
159
Quote requests
1
13
Cost
about $550 a month
nothing
One of the five areas the ads targeted was his own shop. He was paying to show ads to people already in his parking lot.
Sold seven channels, delivering one: about 97% Facebook feed. The games-and-apps placement: zero clicks in a month.
Their clicks fell from 75 to 47 in a month, while they still reported 101% of contracted impressions.
My recommendation: make them fix the targeting or stop it, and put reviews first. Paid search comes later, if at all.
I am not telling you your ads are wasted; I do not know what you are spending or who runs them. I am telling you that is a question worth asking out loud, because it was worth asking here, and nobody had.
What moved on Google
Where he shows up now.
Search, from Klamath Falls
Web result, July
Web result, October 7
windshield replacement
3rd
1st
windshield repair
3rd
1st
auto glass
3rd
1st
rock chip repair
—
1st
oakley sunglasses
not shown
1st
window tint · car window tinting
1st
1st (held)
window tinting · auto window tint
—
1st
home window tinting · commercial window tinting
1st
1st (held)
paint protection film
not shown
2nd
— means not tracked in July. Each October search was run from Klamath Falls with personalization off; the position is epicglass.com among the web results under the map.
“window tint,” October 5: first on the map. On “windshield replacement” he held a steady third, up from not shown at all.“windshield replacement klamath falls,” October 5: his site is the top web link.October 5: Google’s AI answer for “who does windshield camera calibration in klamath falls” named Epic and no other local shop.
Since then. On October 21, from a clean browser with no sign-in, his site was still the first web result on all four searches I read: windshield replacement, auto glass, window tint and car tint. On the map he sat third on all four.
Earlier in October he was first on the map for window tint. He slid to third, behind two shops with “tint” in their business names. I think I know why, and I am working on it. I would rather show you the slide than hide it.
Google’s own AI answer named him and nobody else. ChatGPT has not been as kind; the last time I checked, it still had him behind shops with fewer reviews. Not fixed, and I am not going to pretend it is.
The leads
What the site actually brought in.
Google Analytics, last 90 days. Each key event is a quote request or a tap on his phone number. The line starts when the new site went live.
Quote requests
Taps on his number
First month (Aug 4 – Sep 3)
21
9*
Second month (Sep 4 – Oct 3)
19
19
Since then (Oct 4 – 22)
16
31
Eleven weeks (Aug 4 – Oct 22)
56
59*
*Tap tracking started August 14. Quote requests counted off the actual enquiries, so totals differ slightly from Analytics.
What those requests actually look like
From his quote inbox. Names and phone numbers removed.
Sep 32020 Toyota Camry · windshieldFound him through ChatGPT
Sep 82026 Toyota Camry · full tint“As soon as possible”
Sep 92023 Ford Bronco · full tint“As soon as possible”
Sep 92005 Chevy Silverado · full tint“As soon as possible”
SeptemberA house · 4 windows and a sliding doorResidential tint
Oct 21992 Ford Explorer · windshieldGetting a price
September’s 19: 10 tint, 8 windshield or auto glass, 1 house. 10 of the 19 picked “as soon as possible.”
Google Search Console, August 13, when tracking started, to October 5. Read again on October 22: 223 clicks through October 20.What people searched to find him.
What 20 a month would be worth in your shop
$2,400a month, at your numbers
The only figure here that is mine is the 20 a month: fifty-six quote requests in eleven weeks, rounded down. The job value and the close rate are yours — I am not going to tell you what your work is worth.
Every month
The part that doesn’t show up in a screenshot.
In person: at the shop on August 19, and a monthly meeting since. We go through his numbers and pick the next move.
A written report every month from his own numbers: what came in, where he ranks, what I am doing next.
Ongoing work between meetings: new pages, new articles, fixing listings, answering reviews, and checking where he ranks with the shop open.
Leads from his quote inbox and Google Analytics, August 4 to October 22, 2026. Web results read October 7 from Klamath Falls with personalization off; the map and AI reads are from October 5; the October 21 line is from a clean browser with no sign-in. A “lead” means a quote form sent or a tap on his phone number — not a booked job; only he knows how many of those closed. Rank readings go stale after a fortnight; ask me to run it again. Published with David Pratt’s permission.
