I audited 337 roofer websites. Here's what I found.

I used Google Lighthouse - the same tool Google uses internally to assess websites - to test 337 roofer websites on a simulated mobile phone. I also checked whether each site appears in Google search results, whether it shows up on Google Maps, and how many Google reviews the business has.

This page is the full breakdown: every number, what it means, and what you can do about it. If any term is unfamiliar, the FAQ at the bottom explains them in plain English.

How fast are roofer websites?

Google scores every website out of 100 for speed on mobile. 90 or above is considered good. Below 50 means there are serious problems.

The average mobile speed score across all 337 roofer websites is 61.3 out of 100. Only 14 sites (4.2%) scored 90 or above. More than half sit in the 50-69 range that Google marks as “needs improvement.”

The good news: the average SEO score is 91.3/100. Most sites have the basics right - page titles, descriptions for Google, and mobile-friendly layouts. The speed category is where things break down.

To put it in perspective: if you walked into a roofing trade show and picked a company at random, their website would most likely score between 50 and 69. That's not broken - but it's slow enough that Google considers it a problem, and slow enough that visitors on a phone are waiting noticeably longer than they should be.

Do they pass Google's speed test?

Google measures three specific things about how a page loads, called Core Web Vitals. These directly affect where your site appears in search results. If your site fails them, Google will rank it below a competitor's site that passes - all else being equal.

Only 8 out of 337 sites (2.4%) pass all three.

How long until the page appears?

The most important measurement is how long a visitor waits before they can see the main content of the page. Google calls this LCP (Largest Contentful Paint). Under 2.5 seconds is good. Over 4 seconds is poor.

86.6% of sites take more than 4 seconds before a visitor on a mobile phone can see anything useful. Google's own research shows that 53% of mobile visitors leave if a page takes longer than 3 seconds. The median across all 337 sites is 8.36 seconds - more than three times what Google considers acceptable.

The worst sites in the dataset took over 20 seconds. On a mobile connection, that page will never meaningfully load. A potential customer will have closed the tab, tapped the next result, and called your competitor before your homepage finishes appearing.

What's making them slow

Google Lighthouse doesn't just give a score - it identifies the specific reasons a page is slow and estimates how many seconds each issue costs. The same problems came up across nearly every site I tested.

Images not in modern formats is the single biggest problem - and the easiest to fix. 84.6% of sites are serving images as JPEGs when they could be WebP or AVIF, which are smaller files at the same quality. Fixing this alone would save nearly 4 seconds of load time on the typical site. Every major browser has supported WebP since 2020.

Unused code affects three quarters of sites. This happens when a website builder or plugin loads its entire feature set on every page, even if the page doesn't use those features. Your contact page doesn't need a photo gallery's code. Your homepage doesn't need a shopping cart. But most WordPress plugins load everything everywhere, just in case.

Render-blocking resources affect 80.7% of sites. These are files that the browser has to download and fully read before it can show the visitor anything at all - the digital equivalent of making a customer wait in reception while you read the entire employee handbook before letting them through the door.

Page weight

The total weight of a page - all the images, code, fonts, and files it loads - has the strongest link to speed I found in this data. 67 sites (19.9%) load more than 5MB of data just to show a homepage. That's more data than a 5-minute podcast episode, downloaded every time someone opens the page. Every one of those sites scored below 50 on speed.

The worst example in the entire dataset: one Squarespace site transfers 58MB of data to load a single page. A visitor on a typical 4G connection would wait over 14 seconds before seeing anything. The site scored 18/100. The lightest sites in the dataset - under 500KB - all scored 85 or above. The connection is direct: less data to download means a faster page.

WordPress vs everything else

6 in 10 roofer websites run WordPress (196 out of 337). The rest use a mix of other tools, hand-coded HTML, or platforms I couldn't identify from the outside.

WordPress itself isn't slow - but the way most roofer sites use WordPress is. Themes with dozens of built-in features, multiple plugins adding their own code, unoptimised images uploaded directly from a phone - it all adds up. A WordPress site built properly can score 100/100 (the top site in this dataset proves that). But a WordPress site built the way most web designers build them will land between 50 and 65.

This isn't a reason to avoid WordPress. It's a reason to care about how your WordPress site is built and who builds it.

Who shows up on Google

I searched for “roofer” and “roofing” in 56 different towns across the South-East, recording which sites appeared in the results. This is different from searching for your own company name - of course you'll find yourself if you Google your business. These searches simulate what your customers are typing.

More than a third of all sites - 117 out of 337 - did not appear in any of the searches I ran. These businesses have a website, but when a potential customer in their area searches for a roofer, they won't find it.

On mobile, Google shows a map at the top of most local searches with 3 pinned businesses. Only 74 sites appear in those top 3 map positions for any of the keywords I searched. If you're not one of them, the customer has to scroll past the map, past the ads, and into the regular results before they find you.

Google reviews and the map

Your Google Business Profile is the listing that appears on Google Maps and in the local map pack at the top of search results. For a local trade like roofing, it's arguably more important than the website itself - it's the first thing most customers see, and it's where your reviews live.

The average star rating is 4.81 - high, but that's because satisfied customers are the ones most likely to leave a review. The average review count of 31 hides a wide range: the most-reviewed roofers had 100 to 166 reviews, while many had fewer than 10.

The 82 businesses with no detectable Google profile are invisible on the map entirely. When a customer searches “roofer near me” on their phone, they see a map with 3 to 5 pinned results. If you're not on that map, you're not being considered - no matter how good your website is.

Does a faster site mean a higher ranking?

This is the question most roofers want answered, and the honest answer from this data is: not on its own.

I compared every site's speed score against its Google ranking position. The relationship is weak. A faster site does tend to rank slightly higher, but speed alone doesn't predict whether a roofer will appear on page one.

What does predict map pack position more reliably is Google reviews. Sites with more reviews appeared in the map pack more consistently than sites with faster speeds.

Why do slow sites rank? Because their competitors are equally slow. The average speed score across the entire dataset is 61.3. When every roofer in a town has a 55/100 site, Google ranks based on other signals - reviews, how long the business has been listed, backlinks, how well the Google Business Profile is filled out. Speed becomes a tiebreaker, not the decider.

That said, speed still matters for a different reason: keeping visitors once they arrive. A customer who finds your site on Google but waits 15 seconds for it to load will tap the back button and call the next roofer. You've paid for the ranking in time and effort, then lost the enquiry to a slow page. Fixing your speed doesn't just improve your ranking - it means the people who do find you actually stay long enough to call.

What has the biggest effect on speed?

The strongest link in the entire dataset is between page weight and speed score. The heavier the page (more images, more code, more files), the worse the score. Every site over 5MB scored below 50. This is the most directly fixable issue: compress your images, remove plugins you're not using, and the score will improve.

How each area compares

Roofer website quality varies by area. Here's the average speed score for each region in the dataset.

The differences are small - no area is dramatically better or worse than another. This is consistent with the finding that most roofer sites are built with similar tools (WordPress) and similar approaches, regardless of location. The performance problems are structural, not regional.

The best and worst sites

Top performers

These sites scored well on speed and ranked well in search. They prove it's possible - even with WordPress - to have a fast roofer website that Google rewards.

Bottom performers (anonymised)

I've removed the names and links for the worst-performing sites - the data is real, but there's no value in singling out specific businesses. The pattern is more useful than the names: several of these sites still rank well, carried entirely by their Google reviews.

The most interesting row in that table is the site with 117 reviews ranking #3 organically - despite a speed score of 32/100. That company has built a strong reputation through reviews and Google trusts the signal. But they're leaving money on the table: every visitor who arrives and waits 16 seconds for the page to load is a potential customer lost to the back button.

What does a well-built site look like in this data?

For comparison, the sites I build for tradespeople score 95-100 on Lighthouse across all four categories, load in under 2 seconds on mobile, and pass all three Core Web Vitals. That puts them ahead of every site in this dataset except the very top performers - and unlike most of those, they're built to stay fast as content is added, not just fast on launch day.

That's not magic - it's the result of building with performance as a constraint from the start, not a fix applied afterwards. The same WordPress platform that averages 60/100 in this dataset can score 100/100 when it's built properly.

What you should do about it

If you're a roofer reading this and wondering where to start, these five things will put you ahead of the majority of competitors in this dataset:

1. Set up a Google Business Profile if you don't have one.24.3% of sites in this data had no profile at all. Without one, you can't appear on Google Maps. It's free and takes about 30 minutes. Fill in every field - hours, services, photos, description.
2. Ask every happy customer for a Google review. Reviews are the strongest predictor of map pack position in this data. The businesses with 50+ reviews consistently appeared on the map regardless of their website quality. Make it easy - text them a direct link after the job is done.
3. Get your images converted to WebP format. This is the #1 speed issue and affects 84.6% of sites. If your web developer hasn't done this, ask them why - it's been standard practice since 2020. On its own, this can save 4 seconds of load time.
4. Remove plugins and features you're not using. Three quarters of sites load code that's never executed. If you installed a plugin two years ago and forgot about it, it's still slowing down every page. Deactivate and delete anything you don't actively use.
5. Tell Google what you do and where. Most roofer sites don't include structured data - a small block of code that tells Google your business name, location, service area, phone number, and what you do, in a format it can read directly. It takes a developer 10 minutes to add and helps Google understand your site. More on how Google reads your site →

The data shows you don't need a perfect website to compete in local roofing search. You need to be better than average - and average in this sector is low enough that the bar is genuinely achievable.

Frequently asked questions

Related guides