I scanned 44 GTA service business websites

I ran 44 GTA service businesses through my scanner on a Thursday in September, expecting the usual story about broken code. The code was mostly fine. Forty of the 43 sites that loaded name their city in the title or headline, 34 link their Google Business Profile, and every single one lets the AI search crawlers in.

What they're bad at is the part after somebody finds them. Most of these homepages ask for the call before giving a stranger any reason to trust them, and most of them sound like the next result down the list. The proof usually exists. It's just parked near the bottom of the page, below a stack of quote buttons.

So this turned into less of an SEO study than I planned. It's a study of what happens in the ten seconds after the search works.

How did I run this scan?

I pulled the homepages of 44 GTA service businesses across six trades and ran each one through the FOUND Score, the same scanner behind the free audit on this site, on 24 September 2026. One site didn't load that day, so every number below covers 43.

What does the average local business website score?

77.8 out of 100, with a median of 80 and a spread from 35 to 97. Five of the 43 scored 90 or better, 18 landed in the 80s, and 11 came in under 70.

The spread is the part I keep coming back to. These businesses chase the same customers in the same city, and 62 points separate the best homepage from the worst. Very little of that gap is design taste. It's whether the page does its job once somebody arrives.

Split the five categories into the two jobs a homepage has and the pattern gets obvious. Foundations, Understood and Nearby are about getting found: can Google get in, does it know what you are, does it know where you work. Those three average 16.6 out of 20. Offer and Decision are about getting picked once somebody lands, and those two average 14.1.

Decision, the category that asks whether an interested visitor would actually get in touch, is the weakest of the five at 13.5. Foundations is the strongest at 17.4. Put plainly, most of these sites have solved the problem their web designer was paid to solve, and not quite the one the owner has.

For comparison, the two sites I've built for clients in this set score 90 (Magic At My Door, a Toronto princess party company) and 88 (E&M Equipment, which sells and services air compressors in Orlando). Toggle them on above and you'll see where the gap is: both are ahead on every one of the three found categories, and both are well ahead on Offer.

Why do most sites ask for the call before they earn it?

Because the proof gets treated like a footer item instead of an argument. 25 of the 43 sites had proof a skeptic would believe, such as named reviews that describe the actual job, real numbers, credentials or years in business. Only 3 put any of it near the top, before or beside the first big ask.

Flip that around and it's 40 of 43 businesses asking a stranger to phone them before giving that stranger a reason to. Most do it more than once, too. The median homepage here has roughly ten calls to action below the headline, and the reviews tend to sit somewhere under the last of them.

That order matters more than it used to, because buyers check. 97% of people who use AI to find a business say they sometimes double-check its recommendations against real reviews (BrightLocal, 2026). If they have to scroll past three quote buttons to find yours, a fair number probably won't bother.

Almost nobody is missing proof entirely, which is the encouraging part. 29 sites had real reviews on the page, another 10 mentioned reviews without showing many, and only 4 had none at all. Most of these businesses already have the argument. They're just making it in the wrong order.

So the fix is cheap. Take one strong review, the kind that names the job and the neighbourhood, and move it up beside your first button. You're not adding anything to the page, just changing when the visitor sees it.

What's wrong with the offer on most homepages?

It says what the business does and stops there. 38 of 43 headlines named the service clearly, but only 7 gave a reason to pick that business over the next result, like a guarantee, a same-day promise, a price or a specialty.

The strange part is that most of them have a reason. It's usually sitting one line under the headline, or halfway down the page: 25 years in business, a guarantee, licensed and insured, same-day service. The headline itself leans on "trusted," "best" or "reliable," which every competitor also says, so it can't be the thing anyone picks you for.

Audience was the second leak. Only 12 of the 43 spoke to one clear customer all the way down the page. The most common drift was "homeowners and businesses" in the same breath, with nothing separating the two. Cleaners mixed homes, offices and Airbnb hosts, and party entertainers slipped corporate events into a page built for birthday parents.

Then there's order. Only one page in the sample moved the way a buyer thinks: what you do and for whom, why you, proof, how it works, then the ask. Ten jumped around enough that the AI read scored them zero. And pricing was the quiet one. 22 of 43 published any prices at all, and the rest are betting a stranger will call just to find out, which in my experience most won't. There's more on this in what to say on your homepage.

What else is holding these sites back?

Speed on a phone, mostly, plus a gap in what AI assistants can quote. Google calls a page good when its main content appears within 2.5 seconds, and poor past 4 seconds (Google, web.dev). Not one of the 33 sites PageSpeed could measure made the 2.5-second line on its mobile test. Thirty were past 4 seconds.

The median was 9.5 seconds, which is close to four times Google's line. PageSpeed's mobile test is deliberately harsh (it simulates a mid-range phone on a slow connection), so a visitor on good Wi-Fi sees something faster. The order of fast and slow still holds, though, and the slowest site here took 24 seconds to show its main content.

The AI side is in better shape than I expected. None of the 43 blocks the crawlers behind ChatGPT, Perplexity, Claude or Bing. Two build their pages with JavaScript, so those same crawlers get a blank page, and 23 have no plain summary of two or three sentences under the headline, which is the passage an assistant is most likely to lift.

That last one is worth an hour of your time now. 45% of consumers used AI tools to find local business recommendations in the past year, up from 6% the year before (BrightLocal, 2026). A clear sentence about what you do, where and for whom is the cheapest way to be the answer an assistant repeats. If you want to test it, here's how to check whether ChatGPT recommends you.

Which trades are ahead, and which are behind?

Plumbing, by a clear margin. The six plumbing sites averaged 84.3, and two of them scored in the 90s. Landscaping sat at the bottom at 73.6, with the widest spread in the sample, from 35 all the way to 91. Electrical (78.9), HVAC (78.1), cleaning (77.4) and party entertainment (74.4) filled the middle.

The pattern inside each trade is more useful than the ranking. Landscaping sites scored well on Foundations (18.4, second only to plumbing) and worst of anyone on Decision (11.4), so they're technically sound pages that don't close. Cleaning was the opposite of what I'd have guessed: the best in the sample at being understood by Google (17.9) and near the bottom on Offer (13.3). Party entertainers had the least clear offers of all, at 13.0.

Five to nine sites per trade is a small sample, so I'd treat the trade gaps as a direction rather than a verdict. The chosen half trails the found half in every single trade, though, and that part I'd bet on.

What should you fix on your own site first?

The three things with the widest gap between how easy they are and how few sites do them. Here's the full scorecard first, so you can see where the easy wins are.

First, move one real review up beside your first button. Three of 43 sites did this. Pick the review that names the actual job and the area, put it right next to your call or quote button, and leave the rest where they are. Ten minutes, and it's the single biggest gap in this data.

Second, put your reason to choose you in the headline. Only 7 of 43 managed it, and most of the rest already have the reason somewhere lower down. If it's same-day service or a 10-year warranty, that belongs in the first line people read, not the fourth.

Third, write two or three plain sentences under the headline. What you do, where, and for whom. Twenty of 43 had one. It's what a skimming visitor reads, and it's what an AI assistant is most likely to quote when somebody asks it who to call.

What I wouldn't spend the afternoon on is chasing 100. Five of 43 cleared 90 and plenty of the 80s are doing fine, so close the big gaps and go back to running your business. If you have a second afternoon, speed is next, and our guide on whether a slow website is costing you customers covers where to start.

Frequently asked questions

So where should you start?

With the order of your page, not the look of it. Almost every business here got found. What most of them lost was the next ten seconds: a headline that sounds like everyone else's, and proof sitting below the point where people stop scrolling.

None of the fixes in this article need a redesign. Moving a review, rewriting a headline and adding three sentences under it is one afternoon, and it's roughly the difference between the 3 sites that earn the call and the 40 that ask for it before they have.

It's also most of what I do on client builds. The two sites I built in this comparison score 90 and 88, and Magic At My Door's non-branded clicks rose 143% in its first month on the new site while searches for its name stayed flat (early results, in the case study). If you want to see where your own homepage stands, the free audit runs this exact scan, and if the result raises questions, that's a good reason to talk.