Home / Small business website statistics
Every figure traced to its source
Small business website statistics for 2026, every one checked.
There is no official count of US small businesses with a website. The latest survey says 83%, from 406 owners in August 2025. We measured 122 local business websites in October 2026, and traced 22 of the statistics most often repeated about small business websites: only 2 were both traceable and current.
We sell websites to small businesses, so we have an interest in these numbers. That is why every figure here links to its original source, carries its sample and date, and can be checked.
- Every figure links to its original source, not to another statistics page.
- Every figure carries its sample size and the date it was collected.
- A figure we could not trace is labelled as such, not quietly dropped.
What are the key small business website statistics for 2026?
| Statistic | Source, sample, date |
|---|---|
| 45% of US consumers used ChatGPT or another AI tool for local business recommendations, up from 6% a year earlier | BrightLocal, 1,002 US adults, February 2026 |
| Google users clicked a result on 8% of visits with an AI summary, against 15% without one | Pew Research Center, browsing data of 900 US adults, March 2025 |
| 54% of consumers check a business’s website after reading positive reviews of it | BrightLocal, 1,002 US adults, February 2026 |
| 29% of local business websites run no analytics of any kind | Ours: 122 US local business sites, October 2026 |
| 52% of local business domains publish no DMARC email authentication record | Ours: 122 US local business sites, October 2026 |
| 91% of maintained small business homepages have no heading phrased as a customer’s question | Ours: 44 US small business homepages, September 2026 |
| 18% of ordinary local businesses run on a platform whose own terms say the vendor owns the site | Ours: 44 sites crossed with vendor terms, October 2026 |
| The median published website maintenance plan costs $169 a month | Ours: 100 plan tiers from 34 US providers, October 2026 |
| 95.9% of the top million home pages had detectable WCAG accessibility failures | WebAIM Million, 1,000,000 home pages, February 2026 |
| 91% of new WordPress vulnerabilities in 2025 were in plugins | Patchstack, 11,334 new vulnerabilities, 2025 |
What percentage of small businesses have a website?
Nobody knows precisely, and the number most often quoted is not a population figure. The US Census Bureau does not ask the question, and the SBA does not publish an answer. Every “percent with a website” figure in circulation comes from a survey or an audit run by a company that sells to small businesses.
| Figure | What it measures | Source |
|---|---|---|
| 36,207,130 | Small businesses in the US, 99.9% of all businesses. Most have no employees | SBA Office of Advocacy, February 2026 |
| 83% have a website | A one-day online survey of 406 US small business owners | Clutch, fielded August 2025 |
| 16% of those without one cite cost | The same survey, among the roughly 69 owners without a site | Clutch, August 2025 |
| 17% list no website | Businesses with a Google listing in 24 North American cities | RapidWebLaunch audit of 423,009 listings, autumn 2026 |
Sources: SBA Office of Advocacy FAQ 2026; Clutch, State of Small Business Websites 2025; RapidWebLaunch, State of Small Business Web Design.
The 83% is the best recent survey figure, but read it with its sample. Online panels of “owners” lean towards businesses with employees, and most US small businesses have none, so the true share across all 36 million is very likely lower. The 17% is a better measure of something narrower: businesses visible enough to have a Google listing.
What do local business websites actually contain?
These figures come from our own studies of US local service businesses: plumbers, electricians, HVAC firms, roofers, landscapers, dentists, accountants and pest control companies. We read public page source, contact pages, sitemaps and DNS records. We submitted nothing, and the businesses stay anonymous. The method for each study is below.
Leads and tracking
29% of local business websites run no analytics of any kind (35 of 122).
10% still carry only the Universal Analytics tag, which Google stopped processing in July 2023 (12 of 122).
28% have neither analytics nor call tracking, so they cannot tell which part of the site produced a lead (34 of 122).
89% have a phone number on the homepage that dials in one tap (109 of 122).
30% have no contact form on either the homepage or the contact page (37 of 122).
25% offer online booking (30 of 122).
Please Mention Me study, 122 US local business sites across 8 trades and 18 metros, 6 October 2026. More in why is my website not getting leads.
Email authentication
52% of local business domains publish no DMARC record (63 of 122).
80% have no DMARC record or one that only monitors (98 of 122).
16% publish no SPF record at all (19 of 122).
Same study, public DNS lookups. These records decide whether a contact form’s notification email is trusted by the inbox it is sent to, covered in not getting emails from your website contact form.
Size and structure
The median local business website has 37 pages, with the middle half between 20 and 85 (99 sites with a sitemap).
14% have ten pages or fewer (14 of 99).
Same study, pages counted from public XML sitemaps, excluding posts and media. More in how many pages should a small business website have.
Readability for search engines and AI assistants
91% of small business homepages have no heading phrased as a question a customer would ask (40 of 44).
77% have no llms.txt file (34 of 44).
34% have no structured data at all (15 of 44).
61% have at least one image with no alt text (27 of 44).
The median homepage passes 5 of 9 machine readability checks, and 1 in 44 passes all nine.
Only 1 in 44 blocks AI crawlers in robots.txt.
Please Mention Me study, 44 maintained US small business homepages, September 2026. Detail in website maintenance cost and how many small business websites have an llms.txt.
Accessibility
68% of local business homepages fail at least one basic accessibility check that can be proven from the page code alone (83 of 122). Contrast and keyboard use were not tested, so the true share is higher.
52% have a link with no text or name, usually a social icon or a linked logo (64 of 122).
41% of homepages with a form have a field with no label (29 of 70), and 20% label none of their fields, relying on placeholder text alone (14 of 70).
16% switch off pinch to zoom on phones (20 of 122).
13% load an accessibility overlay widget (16 of 122), and 12 of those 16 still fail a basic check in the code underneath. Half of the dental homepages run one (8 of 16).
Please Mention Me study, the same 122 homepages, 14 October 2026, checked for the WCAG failures provable from source: language, title, image alt attributes, form labels, empty links and buttons, and zoom. Detail in ADA website compliance for small businesses.
Software and security
1 in 4 local business WordPress sites that show their version runs one WordPress.org labels insecure (9 of 36), and only 22 of 36 run the latest release.
25% load a jQuery older than 3.5.0, the release that fixed two known cross-site scripting flaws (18 of 73 that expose a version).
Same 122 homepages, 14 October 2026, against WordPress.org’s version status list. Sites that hide their version are not counted. Detail in website maintenance cost.
Ownership and domains
The registrant’s name was hidden in 43 of 43 public domain lookups, so a public WHOIS check no longer tells an owner whose name their domain is in.
The median local business domain is 15 years old, older than any website built on it (43 domains).
18% of local businesses run on a platform whose own terms say the vendor owns the website (8 of 44), including 6 of the 9 accounting firms.
61% of local business domains are registered through GoDaddy (74 of 121), which renews a .com at a listed $22.99 a year.
7 of 11 pay-monthly website providers say the client leaves without a working site on cancellation, and 24 of 30 providers say nothing about whose name the domain is registered in.
Please Mention Me studies: RDAP lookups of 43 domains, and of 121 for the registrar count, and the published terms of 30 US website providers, October 2026. Detail in do I own my website.
What websites cost
66% of nationally ranked US web design agencies will not take a project under $10,000, while 84% of agencies in mid-size cities would take a $5,000 project (87 agencies).
The median published website maintenance plan costs $169 a month, and the cheapest plan at each provider a median of $94 (100 plans, 34 providers).
Only 34 of 100 maintenance plans include hosting, and 25 do not say.
Plans that state edit time include a median of 120 minutes a month, which works out at about $120 per included hour.
A .com domain renews at a median of $20 a year, against a median first year of $11, and renewals range from $9.98 to $38.38 (17 registrars). The wholesale cost is $10.46 including ICANN’s fee, rising to $11.17 on 1 November 2026.
Entry-level shared hosting renews at a median 2.93 times its introductory price (9 hosts).
Home services website subscriptions have a median of $249 a month, and one-time builds a median of $4,650. The three year cost of the cheapest plan varies fifteen-fold between providers (55 plans, 21 providers).
Please Mention Me studies of published prices, September and October 2026. Detail in what a website redesign costs, website maintenance cost, how much a domain name costs and HVAC website design.
Which third-party small business website statistics are worth citing?
These come from primary sources published in 2025 or 2026, each checked on the source page in October 2026.
| Theme | Statistic | Source, sample, date |
|---|---|---|
| AI and local search | 45% used AI tools for local business recommendations, up from 6% a year earlier. A rise that steep may partly reflect a changed question | BrightLocal, 1,002 US adults, Feb 2026 |
| Reviews | 97% read reviews for local businesses, and 54% then check the business’s website | BrightLocal, as above |
| Search clicks | Users clicked a result on 8% of Google visits with an AI summary, against 15% without; 1% clicked a link inside the summary | Pew, 900 US adults, Mar 2025 |
| Search clicks | AI Overviews correlate with a 58% lower click-through rate for the top-ranking page | Ahrefs, 300,000 keywords, Feb 2026. Correlational |
| ChatGPT use | 34% of US adults have used ChatGPT, about double the share in 2023 | Pew, 5,123 US adults, Mar 2025 |
| Mobile | 91% of US adults own a smartphone, and 16% use only a smartphone to get online | Pew, 5,022 US adults, 2025 |
| Accessibility | 95.9% of home pages had detectable WCAG failures, an average of 56.1 errors per page; 83.9% had low contrast text | WebAIM Million, 1,000,000 home pages, Feb 2026 |
| Security | 11,334 new WordPress ecosystem vulnerabilities in 2025, up 42%; 91% in plugins | Patchstack, its vulnerability database, 2025 |
| Security | Ransomware was present in 88% of breaches at small and medium businesses, against 39% at larger organisations | Verizon DBIR 2025, 12,195 confirmed breaches |
| Platforms | WordPress runs 40.1% of all websites, and 58.6% of sites whose content system is known | W3Techs, top 10 million sites, Oct 2026 |
| AI adoption | 19.8% of US businesses used AI in any business function | US Census Bureau, Business Trends and Outlook Survey, May 2026 |
Which small business website statistics are wrong?
Statistics pages copy from each other, and figures drift as they are copied. We took the 22 statistics repeated most often across the pages that rank for small business website statistics, and traced each one back to its original source. Where we could open the original, we did.
| Verdict | Count | What it means |
|---|---|---|
| Verified and current | 1 | Traceable, quoted correctly, recent |
| Superseded | 1 | Correct, but a newer edition of the same source exists |
| Stale | 8 | Traceable and broadly accurate, but years old |
| Misquoted | 6 | The original says something narrower or different |
| Untraceable | 6 | No original source could be found |
Only 2 of the 22 are both traceable and current. The pages repeating them are not named here. The point is the figures, not who copied them.
| As usually repeated | What the original says | Verdict | Use instead |
|---|---|---|---|
| 83% of small businesses have a website, up from 64% in 2018 | Clutch, one-day online survey of 406 US owners, August 2025. The 2018 figure came from a separate survey of 351 owners, so the two are not a measured trend | Verified, with its sample | Quote 83% with its sample, and drop the trend |
| 72% of businesses have a website in 2025 | The figures match a 2021 release: 71%, 500 owners, fieldwork December 2020 | Misquoted: 2021 data, redated | Clutch 2025, 83%, 406 owners |
| 73% have a website | A May 2022 release, 1,003 owners and managers | Stale (2022) | Clutch 2025 |
| 71% globally | The 2021 US survey | Stale, and not global | No global figure exists |
| 26% without a website cite cost | Clutch 2018 | Stale. Clutch’s own 2025 figure is 16% | 16%, Clutch 2025 |
| 94% of first impressions are design related | A 2004 study of 15 UK women judging menopause health sites: 94% of their comments on why they rejected a site concerned design. For sites they trusted, content dominated | Misquoted, 22 years old | Do not use for business websites |
| 75% judge a company’s credibility by its website design | Attributed to a 2002 Stanford study, which reports design look in 46.1% of 2,440 comments. No 75% appears in it, and its categories overlap, so they cannot be added together | Untraceable | Stanford 2002, 46.1%, with its date |
| 38% stop engaging if the layout is unattractive | Adobe survey of 2,008 US adults, September 2015, about content on any device | Stale (2015), and not specific to websites | Quote with its year and scope |
| 57% will not recommend a business with a poorly designed mobile site | No original found. The closest real study, Google 2012, reports different figures | Untraceable | Do not use |
| 61% (or 74%) leave if they cannot find what they want within five seconds | Google 2012, 1,088 US smartphone users: 61% would move on if they did not find what they wanted “right away” on a mobile site | Misquoted, 14 years old | Quote “right away”, mobile, 2012 |
| 81% research online before buying | A 2013 bank survey of 3,220 people who had recently made purchases of $500 or more | Misquoted: major purchases only | BrightLocal 2026 for local search behaviour |
| 84% think a website is more credible than social media alone | Verisign survey of 787 US internet users, 2015 | Stale (2015) | Quote with its year |
| 53% will not wait more than three seconds | Google and DoubleClick 2016: 53% of mobile site visits are abandoned when a page takes over three seconds to load | Stale, and visits, not users | Quote as mobile visits, 2016 |
| Visitors form an opinion in 0.05 seconds | Lindgaard and colleagues, 2006, small laboratory experiments | Stale (2006), accurate | Quote with its year |
| 88% are less likely to return after a bad experience | Probably a 2009 performance monitoring vendor survey, no longer available | Untraceable | Do not use |
| 70% of small business websites have no call to action | A 2013 study of 200 small business-to-business firms, homepage only, excluding phone numbers and contact links | Misquoted, 13 years old | Our data: 4 of 122 local sites lack both a phone link and a form |
| 43% of cyber attacks target small businesses with outdated platforms | Verizon 2019: 43% of breaches involved small business victims. “Outdated platforms” is not in the source | Misquoted, stale | Verizon DBIR 2025: ransomware in 88% of small and medium business breaches |
| 97% search online for local businesses | An older BrightLocal figure. The 2026 survey says 97% read reviews for local businesses, a different claim | Stale, replaced by a different finding | BrightLocal 2026 |
| 99% use the internet to find local businesses | No original found | Untraceable | Do not use |
| Businesses with websites are 2.8 times more likely to grow | No such study found. A 2017 Deloitte report for Google says digitally advanced firms grew revenue nearly four times as fast | Untraceable as stated | Deloitte 2017, with its date and wording |
| The average time on a page is 54 seconds, and video raises it by 88% | The pages repeating these say the original studies are no longer available | Untraceable | Do not use |
| 94.8% of home pages have accessibility issues | WebAIM Million 2025 | Superseded | WebAIM 2026: 95.9% |
Each row was traced to its original where one exists. Original wording and links for every row are in our research notes.
How one statistic changed as it was copied
The clearest example is a 2012 Google survey of 1,088 US smartphone users. It found that 61% would quickly move to another site if they did not find what they were looking for “right away” on a mobile site. As it has been copied from page to page, three things changed:
- “Right away” became “within five seconds”.
- 61% became 74%, apparently borrowed from a separate finding in the same study.
- “Find what they were looking for” became “takes longer than five seconds to load”, which is how a live AI summary in Google search phrased it in October 2026.
The finished version is a claim about page speed that the original never made, attributed to a fourteen year old study about something else. That is why every figure on this page carries its sample and its date.
None of this means websites do not matter. It means the case for them was being made with numbers nobody had checked. The current evidence, the reviews survey, the clicks data and our own counts, makes the case better than the old figures did, and it has the advantage of being true.
If you write about small business websites, take any figure from this page. Just take the sample and the date with it.
How were these statistics measured?
| Study | Sample and method | Limits |
|---|---|---|
| 122 local business websites, October 2026 | 8 trades, 18 US mid-size metros, picked from organic results excluding directories and national chains. One automated request per homepage and contact page, the XML sitemap, and public DNS. DMARC is a TXT lookup on the domain’s DMARC record | Sites behind bot protection are under-represented: 41 of 131 candidates blocked the request and were replaced. DKIM is checked on common selectors only, so it is a floor. Tag Manager may or may not contain analytics |
| 44 small business homepages, September 2026 | The same frame, scored on nine machine readability checks | These 44 are part of the 122. Homepage only |
| 43 domain registrations, October 2026 | Public RDAP lookups for the 44 domains; one returned no record | Shows the registrar, not the account holder |
| Accessibility, software and registrars, 14 October 2026 | The same 122 homepages re-read as raw HTML, before scripts run, for the WCAG failures provable from code; WordPress and jQuery versions where exposed; RDAP for the registrar of 121 domains | A floor for accessibility: contrast, keyboard use and anything built by scripts were not tested. Hidden software versions are not counted either way |
| Terms of 30 website providers, October 2026 | Published terms, service agreements and FAQs, coded clause by clause, with quotes checked word for word | Terms change. Silence was coded as silence |
| Prices: 87 agencies, 100 maintenance plans, 55 trade web plans, 20 domain registrars, 13 hosts | Clutch profiles, and prices read on each provider’s own page | Published prices lean towards the budget end; agencies that publish no price are not in the plan samples |
Third-party figures were included only if the original source could be opened, was published in 2025 or 2026, and states a sample or method. We sell websites to small businesses, and these studies were run by us on public data; the definitions above are enough to repeat any of them.
How should I cite these statistics?
Quote any figure with a link to this page or to the statistic’s own anchor. Each of our statistics has one: for example, pleasementionme.com/small-business-website-statistics/#stat-dmarc. A suggested credit line: “Please Mention Me, Small business website statistics 2026”. For third-party figures, cite the original source linked in the table rather than this page.
Changelog. 13 October 2026: page published, with October 2026 data. 14 October 2026: added accessibility, software, registrar and domain price figures. Figures will be updated as new editions of the third-party sources appear.
Common questions
What percentage of small businesses have a website?
There is no official figure. The most recent survey, by Clutch, says 83%, from 406 US small business owners in August 2025. Online panels lean towards businesses with employees, and most US small businesses have none, so the true share is very likely lower.
How many small businesses do not have a website?
No reliable count exists. Clutch’s 2025 survey puts it at 17% of the owners it asked, and a 2026 audit of businesses with a Google listing found 17% listing no website. With about 36 million US small businesses, both figures describe a subset, not the whole population.
Do customers still check a business’s website?
Yes. BrightLocal’s 2026 survey of 1,002 US adults found 54% check the business’s website after reading positive reviews. A website is often the second stop, after the reviews, before the call.
Are AI assistants changing how customers find local businesses?
The data points that way. BrightLocal reports 45% of consumers used AI tools for local recommendations in 2026, up from 6%, though the jump may partly reflect a changed question. Pew found Google users click results less often when an AI summary appears.
How much does a small business website cost?
In our price studies, a five to eight page site has a median published one-time price of $2,499, home services specialist builds a median of $4,650, and maintenance plans a median of $169 a month. Many nationally ranked agencies will not start under $10,000.
Which small business website statistics are out of date?
Many of the most repeated ones. Of 22 we traced, only 2 were both traceable and current. “94% of first impressions are design related” comes from a 2004 study of 15 women judging health sites, and “75% judge credibility by design” could not be traced to any study at all.
See where your own site stands
Our free checker runs the nine machine readability checks behind several of the figures above on your homepage. About ten seconds, no email required.