US Small Business Websites 2026.

Four states make up nearly half the named locations.

Each business's market is written down in our research notes, usually as a city and state. California came first with 94 businesses, then New York, Florida and Texas. Between them those four states hold 94, 54, 51 and 41 businesses, close to half of the businesses with a named state.

101 businesses (16.9%) did not name one place. Most of these sell across the country or online: makers who ship to every state, coaches and consultants who work by video, musicians and creators whose audience is national.

The spread follows population and business density more than anything about websites. It also follows who answered our outreach, which is why we describe the sample rather than claim it speaks for every American small business.

Most American homepages ask visitors to go looking.

On 62.0% of the 400 US homepages there was no form a visitor could fill in. On 52.2% the phone number, where there was one, did not dial when tapped. 37.5% had neither. Almost all of them did link to a contact or booking page somewhere else, so the business was reachable; the first screen just did not offer the way in.

The state groups below are small, so read them as a picture rather than a ranking. Still, the spread is wide. Florida homepages were the most likely to carry a tap-to-call link, while New York homepages were the least likely, and New York also had the most homepages that were not set up for phones.

Only states with at least 30 homepages measured are shown.

Squarespace is far more common in the US.

WordPress still ran the most American homepages, at 34.5%, a little below its share across the whole study (38.8%). The difference is Squarespace: 12.8% of US homepages, compared with 6.9% in the UK and 5.8% in Canada.

GoHighLevel, the marketing system many US agencies resell to local businesses, showed up on 2.8% of US homepages, against 0.5% in the UK. A third of homepages (32.8%) showed no platform signature at all: hand-coded sites, frameworks such as Next.js and newer builders that leave no trace in the page.

For an owner the platform matters less than what it costs to change things. On a closed builder, leaving means rebuilding the site; on self-hosted WordPress, it often means a careful move where the business email may live on the same hosting.

More generated imagery than anywhere else in the study.

Our written notes on 509 American sites describe stock or stock-looking photos in key places on 39.1% of them. That is close to the rest of the study. What stands out is generated imagery: 10.8% of US sites, against 7.3% in the UK and 7.9% in Canada.

14.1% of US sites still showed placeholder or demo content, such as sample testimonials or empty template sections. 21.2% were built with a WordPress page builder such as Elementor or Divi.

The other side of the same notes is encouraging. 89.9% of the 597 American businesses had real photos or films of their own somewhere online, which could replace those stock and generated images today.

Instagram and YouTube hold what the website leaves out.

When we research a business, we note where its real photos, films and reviews are. For American businesses that was Instagram most of all (44.4% of the US notes), then YouTube (33.7%) and Google reviews (26.6%). YouTube is noticeably stronger in the US than in the UK, often as a channel of job walk-throughs, client stories or show episodes.

WhatsApp is the reverse. It appeared in 0.7% of US notes and on 3.2% of US homepages, where in the UK it was on 21.1%. American customers call or text, so for US sites the tap-to-call link and a text button do the work that a WhatsApp button does elsewhere.

These counts come from written notes, so they are a floor: a note that does not mention YouTube does not prove the business has no channel.

The heaviest pages are heavier in the US.

The typical US homepage was in line with everyone else: a median of 2.18 MB against 2.27 MB for the whole study. The difference is at the top. One in ten US homepages (10.0%) downloaded more than 10 MB on its first load as a phone, and the heaviest tenth started at 9.8 MB. In the UK that line was 7.4 MB.

Pages that heavy usually carry an autoplaying background video, full-size photos straight from a camera, or several tracking and chat scripts. On a weak signal in a parking lot they take a long time to appear.

Layout on phones was mostly fine: 9.2% of US homepages lacked a mobile viewport or scrolled sideways, the lowest share of the three countries with their own edition.

A short check for an American small business site.

Open your homepage on your phone, on cell data rather than Wi-Fi, and go through these. Each one comes from something we saw on real US sites this autumn, and none of them needs a new website.

How the US numbers were made.

This edition uses exactly the method of the main report, limited to businesses whose market is in the United States. A business counts once. Its state is read from the market written in our research notes; where the notes name two states, the first one counts.

486 US business domains were opened once on 1 November 2026 in headless Chrome set up as a phone (390 by 844 pixels, touch, iPhone browser identity). 400 loaded fully and make up the homepage figures. 17 did not load at all, 6 showed a parked or "coming soon" page and 63 blocked automated visits or did not finish loading, and these are left out of the percentages.

The sample is small businesses that replied to Astra Labs about their website, so it leans toward owners already thinking about a new site. Our own paying clients' sites are excluded. We never publish which business had which result.