Contractor Website Speed: We Timed 18 Local Sites

Contractor website speed is mostly a phone problem, and most local sites fail it. We loaded 18 Sacramento-region contractor websites the way a mid-range phone on a slow 4G connection would, three times each, on November 4, 2026. The median site took 7.0 seconds to show its main content. Google counts 2.5 seconds or less as good. One of the 18 cleared that line, and five took longer than ten seconds.

The short version:

We build websites for contractors and our own homepage is in the test, so read this with that in mind. Everything here is measured, the method is at the end, and the one site that passed is named. The rest are not, because the pattern is the point.

Why the phone is the test that matters

A homeowner with a burst pipe or a dead air conditioner is not at a desk. They search from the driveway, from the couch, or from the hallway with a bucket in the other hand. They tap one result, and if the page is still blank they tap the next one.

Google's own position is measured. Its page experience guidance (last updated September 22, 2026) says its "core ranking systems look to reward content that provides a good page experience," and also warns that good scores do not "guarantee that your pages will rank at the top." Its Search documentation sets the target we used: the main content should appear within 2.5 seconds, and the first thing on screen within 1.8 seconds. Google runs the search engine it is describing, so treat those as its stated aims, not as a promise.

For a sense of where 7.0 seconds sits: HTTP Archive's 2025 Web Almanac, which uses real Chrome visitors' data from July 2025 across the web, found that 62% of websites load their main content within 2.5 seconds on phones. In our group of 18 local contractor sites it was 1 in 18, about 6%.

We found no study of how speed changes calls to a contractor, and we are not going to borrow a retail one. Old figures about shoppers leaving after three seconds come from 2016 and 2017 research on online stores. What we can show is how long a phone waits.

The one that passes

A1 Plumbing and Drains shows its main content in 1.8 seconds, and the whole page weighs 756 KB. It is a plain site: a few images, a short page, and nothing that makes the phone wait for something heavy first. A builder-made site can pass. It does not need to be elaborate, it needs to be light.

Four more came close, showing their main content between 2.9 and 3.7 seconds, slower than Google's line but within about 1.2 seconds of it. One of those four loads a stock video file and still manages 3.1 seconds, so a video is not always the whole story.

What the other 17 make a phone download

We read each test report's three largest downloads. These are what we saw loading on the slow sites, not a proven cause for each one, and they were ordinary.

• One oversized picture. The site that took 16.0 seconds leads with a 1.6 MB background picture, uploaded in 2019. Another downloads a single 3 MB photo, and its screen stayed blank for 4.3 seconds and then showed everything at once. A third takes 13.5 seconds and downloads a 753 KB hero picture plus its form checks. A phone screen needs a fraction of that picture weight.
• Photos that were never shrunk. The slowest site in the group, at 23.0 seconds, downloads 8.4 MB, including three uploaded files of 0.5 to 1.9 MB each.
• Video. Two homepages load a video file. One is 27 MB on its own, and that homepage downloads 30.9 MB in total, twelve times the weight of a typical phone home page. It showed its main content in 11.2 seconds on the test. The video plays behind a headline, which looks fine on an office monitor and costs a phone a very large download for something nobody asked for.
• Tools the page has to fetch first. On three sites, Google's form-checking script (the one that blocks spam from the contact form) is among the three biggest downloads, up to 839 KB. On another, a chat widget script is a 1.1 MB download. These tools are useful. They should load after the page, not before it.
• Pages that jump. Six of the 18 shift visibly as they load, past Google's 0.1 line for layout movement, and two score at least ten times that. On a phone that is how you tap the wrong button.

Weight is not the whole story. The 4.9 MB site loaded its main content in 3.1 seconds, and a 2.3 MB site took 9.2. What matters is what the page makes the visitor wait for first. Even so, eight of the 18 weigh more than 2.6 MB, the median phone home page in the Web Almanac's page weight chapter.

Our own homepage

We ran www.flarebuilt.com through the same test as a control. It showed its first content in 1.2 seconds and its main content in 2.68 seconds, at 645 KB, about a quarter of the group median. That is 0.18 seconds past Google's line, so by this test we do not pass either. It is the same result as the 2.76 seconds we reported when we timed 43 brokerage websites. The cause is the script behind the homepage's interactive parts, not pictures or video, and we are still working to cut it.

A five-minute contractor website speed check

1. Run your homepage through PageSpeed Insights and read the mobile result. The number labelled Largest Contentful Paint is the one in this article: when the main content appears.
2. Check your biggest picture. A hero photo over 500 KB is almost always larger than a phone needs, and a photo straight off a phone camera can be 3 MB or more.
3. Look for a video on your homepage. If there is one, check its size. Anything over a few megabytes on a phone is a cost you are choosing.
4. Ask what loads before the page. A chat widget, a booking tool or a spam check on the contact form should wait until after your headline and phone number are on screen.
5. Open your site on your own phone, on cellular data, not Wi-Fi. Count how long it is before you could read it or tap the call button. Our click-to-call check covers what to do once the page is there.

What we would build

This is the part we sell, so read it as a specification you can hold anyone to, including us.

A homepage that shows its headline, your phone number and a job photo inside Google's 2.5 seconds on a mid-range phone. That means photographs sized for the screen they land on (see what makes contractor website photos work), no autoplay video on phones, and chat and booking tools that load after the page, not before it.

A Launch build, a sharp one-page site, starts at $900 and is typically live in about 48 hours. A Signature build starts at $4,500. What each includes is in our contractor website cost guide.

If you want this test run on your site, tell us about your business, or build your package in about two minutes.

How we tested

The sample is the 18 websites from our original audit of licensed contractors: plumbing, heating and air, electrical, roofing, low voltage and general building, in Sacramento, Placer, El Dorado and Yolo counties. They are the same 18 as our other contractor checks, and one of the 18 addresses is an inner page of the company's site, tested as listed. Our own homepage was added as a labelled control.

Each page was loaded three times on November 4, 2026 with the open-source test engine behind Google's PageSpeed Insights, version 13.5.0, at its default phone settings: a simulated mid-range phone with a processor four times slower than a laptop on a 1.6 Mbps connection with 150 milliseconds of delay. We report the median of the three runs. Every run identified itself as an ordinary Android phone browser, because testing tools that announce themselves get turned away by some sites' bot protection, and a blocked test looks like a slow site. All 19 pages produced three valid runs, so none had to be set aside.

This is a laboratory test on one simulated device and connection, not a measurement of real visitors, and times vary by hour and by network. We did not measure whether a slow site loses a contractor any calls. What the test does show, reliably, is which sites make a phone wait, and for what. The Web Almanac comparison uses a different method (real Chrome visitors, across all kinds of websites), so treat 62% versus 6% as a sense of scale, not a like-for-like score.