Page one or page two? 122 roofing company websites, measured side by side

Ask a roofer stuck on page two what the companies on page one have, and the answer is usually some version of more: more pages, more blog posts, more homepage copy, a chat widget, financing. It is also the pitch most roofers hear from whoever sells them content. It is rarely checked, because studies of what ranks, including our own HVAC study, measure only businesses already on page one. You cannot learn what separates page one from page two without looking at page two.

So we looked at both. On September 16, 2026 we captured the first two pages of Google for “roofing company” in 11 American metros, screened out everything that was not a roofing company, and measured the website of every roofer that remained: 122 companies, 48 of them on page one and 74 on page two.

The belief did not hold up. On every website measure we took, the typical page-one and page-two roofer look nearly the same. What differs is how many roofers each page holds.

The benchmark: 152 pages, 1 blog post, 3 of 8 features

The numbers to compare your own site against:

Measure	Page one (1 to 10)	Page two (11 to 20)
Roofing companies measured	48	74
Median pages in sitemap	181	148
Median blog pages	1	1
Share with no blog pages at all	38 percent	31 percent
Median blog posts updated in 12 months	2	2
Median homepage words	1,367	1,345
Median features, out of 8	3	3
Median unverifiable superlatives on homepage	6	5
Median checkable numeric claims on homepage	2	2

Pooled, the median ranking roofer has 152 pages, 1 blog page and 1,345 words on its homepage. Sixty-two percent have fewer than 200 pages and 37 percent have fewer than 100. In that range, you are not behind the field.

Site size: a gap in the tail, not in the middle

The one measure where page one leads is site size. The median page-one roofer has 181 pages against 148 on page two, the upper quartile 337 against 225, and the mean 381 against 200. The gap lives mostly in a handful of very large page-one sites, not in the typical one.

As a rank test the difference is not significant: a correlation of 0.089 across 103 sites with a readable sitemap. That is not proof of no effect; a gap this small is beyond what 103 sites can separate from chance. Call it a lean. Page one had the larger median site in 8 of the 11 markets.

Two details push against “bigger wins.” Roofers in positions 1 to 3 have the smallest median site of any band, 132 pages, against 319 for positions 7 to 10. Only 14 roofers sit in the top three once directories are removed, a small group, but the opposite of what the belief predicts. And the blog makes no visible difference: the median is 1 blog page on both pages, and 28 of the 71 sites linking to a blog from the homepage have 1 blog page or none in their sitemap.

Features: no difference anywhere

We scored every readable homepage on the eight features we used for med spas and veterinarians: online booking, a form, chat, financing, a pricing link, review schema, FAQ schema and a blog link. The median is 3 of 8 on page one and 3 of 8 on page two. Not one of the 108 sites we could score carried more than 6.

None of the thirteen homepage flags we read differed significantly between the pages. Online booking appears on 5 percent of page-one sites and 9 percent of page-two sites; financing on 62 and 67 percent. The largest gap was a mention of emergency or same-day service, 45 percent against 29 percent, and even that the sample cannot confirm.

One measure did move with position. Across positions 1 to 20, roofers higher on the page make slightly more checkable numeric claims on their homepage, such as years in business or roofs completed: a rank correlation of 0.234 across 108 homepages. It is weak, and one of eight position tests, so one clearing the bar by chance would not be surprising. Treat it as a lead. It is cheap to act on: 20 percent of ranking roofers make no numeric claim at all.

What does differ: the page itself

Here is what the belief misses: the two pages do not hold the same number of roofers.

	Page one	Page two
Results captured	95	105
Directories, marketplaces and other non-roofers	43	28
Share that is not a roofing company	45 percent	27 percent
Median roofing companies per market	5	7

In a typical metro, 5 roofing companies appear on page one. In Nashville and Salt Lake City, only 2 did. The rest of page one belonged to Yelp, the Better Business Bureau, a shingle manufacturer’s contractor locator, HomeAdvisor, Reddit and a home improvement retailer. The manufacturer’s locator was on page one in 7 of the 11 markets. Yelp held a top-three position in 8 of them.

Two consequences follow. A roofer on page two is competing for about five real slots, not ten. And several of the results above it are places a roofer can appear inside: a directory profile, a manufacturer’s certified contractor listing, a review site. That is a second route onto page one, and it does not run through the website.

Some roofers take the other route directly. In 6 of the 11 markets, page one included a roofing company’s own blog post ranking “the best roofing companies” in that city, and in Kansas City one held position 1. We kept those companies: the subject is the business, not the page that ranked.

What verification changed

Our first capture lost 8 of its 22 searches to a 12-second time limit shorter than the search provider needed. Five markets were missing page one, which would have compared page one from six cities against page two from nine. The searches had completed on the provider’s side, so repeating them recovered all 22 from its cache without spending another. The engine now waits longer for searches.

The first complete run then kept 11 sources that are not roofing companies, among them a contractor verification directory, a media ranking list, a trade association’s contractor finder, Yahoo Local, a classifieds site, a home improvement retailer and a roofing franchise’s national site. One reported more than a million pages in its sitemap and dragged the average far above anything a roofer runs. With them removed, the pooled mean is 269 pages against a median of 152.

Google’s page two also repeated 5 results from page one of the same search. Those are not second slots, so they are removed before screening, which is why the sample counts 200 results rather than 205.

Limits: what this cannot tell you

This study compares page one with page two. It cannot say what separates either from roofers outside the top 20, who are not in the data.

Every measure here is on the website. We did not measure backlinks, domain age, Google Business Profile strength, reviews or proximity to the searcher. Those are likely explanations for the ordering, and websites that look alike on both pages are consistent with them doing most of the work.

Results were captured once, from one location setting per market, for one query. Rankings vary by device, location and time. Page counts come from published sitemaps, and 19 roofers had none we could read, so they are missing rather than zero. Blog volume counts sitemap URLs under a blog, news, articles or resources path, and a site that publishes posts at the root of its domain is undercounted. Homepage features are read from the homepage only; 11 homepages returned too little readable text to measure and 3 did not respond, so they are reported as unmeasured. A company ranking in more than one market is counted once. The 71 excluded results are in the dataset with their reasons.

The number to write down

If your roofing company is on page two, the median site on page one is 181 pages with 1 blog page and 3 of 8 conversion features, and the median site next to yours on page two is 148 pages with the same blog and the same features. The distance between you is probably not on your website. Count how many of the results above you are actually roofers, and look at which of the rest you could appear inside.

The full dataset, with every exclusion and its reason, is below.