Case Study: Tint Guy LLC Cut Cost Per Lead 58% in Three Months

Tint Guy LLC, an automotive window tint shop in Murfreesboro, Tennessee, was spending about $1,700 a month on Google Ads across six campaigns, one of which was paying $154 for a lead and another $45 to buy its own business name. We rebuilt the search side around a single campaign. Three months in, cost per lead was down 58% on the same budget. We also checked whether the account was counting the same phone call twice before changing anything, and the answer was not the one we expected, which is the more useful part of this page.

How we measured, and why it is the first thing on this page

A lead here means one thing: a form submitted on the website, recorded server-side, counted once per click. It is not a phone-link tap, not a modeled store visit, not a page view. We picked the strictest available number on purpose, and the reason is the section further down.

All three months use the same campaign, the same conversion action and the same counting rules, so the trend is not an artifact of the measurement changing underneath it. Every figure came from the Google Ads API. Nothing is estimated or projected.

What this is not: a year-over-year comparison. The campaign launched in month one, so this is an improvement curve within a single campaign over three months, not proof against a prior year.

What the account looked like before

Six campaigns, five of them run by a third-party automated tool, each capped at roughly $150 a month. Their cost per lead over the prior year:

Two of those were selling services the owner had specifically told us were out of scope. The commercial campaign alone was past the point where we shut a campaign off, every month, for a year.

What we did instead

One search campaign, automotive tint only, 21 keywords and 35 negatives, geo-fenced to a 30 mile radius of the shop. Server-side conversion tracking so a lead is recorded when it actually lands rather than when a browser fires a tag. Then three months of ordinary work: reading search terms, adding negatives, and letting the bidding learn on a clean signal.

Cost per click fell from $2.18 to $1.30 over the same period, and clicks rose from 248 to 429. Better keyword and negative coverage means a higher share of the auctions entered are ones worth winning, and quality signals follow.

If you run a tint shop, the harder question is which jobs are worth advertising at all. We cover that, job type by job type, on our window tint marketing page.

What we found in the account, and what we got wrong about it

This is the section most case studies leave out, which is exactly why it is here.

While auditing the conversion setup we found the account had eight separate phone-call conversion actions, all counted in the same Conversions total. Three of them recorded the same event type: a call placed from an ad. The largest was named after a Black Friday promotion and had been firing every month since February.

Our first read was that one phone call was being counted twice, which would have made the account’s cost per conversion look about a quarter better than it really was. We wrote that down, with a prediction, and prepared the fix.

Then we tested it against Google’s own call records before touching anything. Over two weeks the campaign received 48 ad calls lasting a minute or longer, and the two ad-call actions counted 50 between them. They were splitting real calls, not duplicating them. Removing one would have deleted about 35 real calls a month from the signal the bidding learns from. We graded our own theory wrong and left both actions in place.

What we did change: a tap on the phone number on the website no longer counts as a lead, because the call it starts is already counted when it lasts 60 seconds or more.

This is why the number at the top of this page is the strict one, and why we test before we fix. A tracking change made on a theory would have made the reports look tidier and the campaign worse.

What happened next

The campaign kept running, the website was rebuilt around what the shop actually sells (new pages for paint protection film and ceramic coating, with Tint Guy LLC’s Ceramic Pro Elite products named), and we measured everything weekly. Where things stood at the end of September 2026:

The map pack is measured from 49 points in a grid across town every week, for “window tint near me”, “car window tinting” and the city search, because one ranking number means nothing on Google Maps: it changes with where the searcher stands. The grid also shows where the shop still loses, the north side of town, where a closer competitor wins the search. That is the next piece of work.

Not just the ads: the rest of the work

The ads are the easiest part to measure, so they lead this page. They were one piece of a full engagement that also covered the website, Google search, the Google Business Profile and AI answers, because a click only turns into a job if everything behind it holds up.

The AI side has its own write-up: one week of AEO work, then a customer’s message.

What we would not claim

Three months is three months. Some of the improvement is a new campaign’s bidding settling down, which happens without anybody being clever, and we would expect the curve to flatten. Leads are form submissions, not booked installs. We can see that someone asked to be contacted; we cannot see who ended up in the bay, and we will not put a revenue figure in front of anyone that we cannot trace.

What does a marketing agency actually do for a window tint shop?

Tristar Marketing Solutions is a marketing agency for window tint, PPF and ceramic coating shops that makes a shop findable in the three places customers actually look: Google search, the map pack, and AI answers. On the paid side for a tint shop that means one campaign aimed at the service the shop actually wants to sell, negatives worked weekly, conversion tracking that counts a lead once, and a cost per lead the owner can check against what a job is worth.

Every figure pulled live from the Google Ads API. A lead means a server-side form submission counted once per click, deliberately excluding phone-link taps, modeled store visits and page views. All three months use the same campaign, conversion action and counting rules. The September figures come from the same Google Ads account (calls from Google’s call records, 60 seconds or longer), the live Google Business Profile, and a 49-point weekly map-pack grid. Tint Guy LLC has been a Tristar client since September 2023 and is named with its permission. Results describe one business over one specific period and are not a prediction of what any other business will see.

More client results

• Tint Guy LLC: 900+ reviews and a zero percent AI citation rate
• Marine dealer: average position 48 to 10 in one season
• Towing company: page two to page one in twelve days