Owning your code, platform lock-in, builder speed
You cannot take a Wix site anywhere else; it only runs on Wix's servers.
Wix Help Center, Exporting or Embedding Your Wix Site Elsewhere · undated page, seen 2026 · the vendor’s own data · support.wix.com ↗
The sentence in the source
“In order for your site to work properly, it needs to be hosted and operated on Wix's servers. … The reason you can't use another host for your Wix site is that the SaaS architecture does not support external hosting. … Since Wix is a SaaS solution, your site must run on Wix's servers.”
Where: Body text, first section (undated page, fetched 2026-10-16)
Confidence: Verified verbatim
Wix says the content is yours, but only the domain can be managed elsewhere, not the site.
Wix Help Center, Exporting or Embedding Your Wix Site Elsewhere · undated page, seen 2026 · the vendor’s own data · support.wix.com ↗
The sentence in the source
“You can, however, connect a domain to your Wix site, and purchase and manage the domain with another provider. … the content you build on Wix belongs to you.”
Where: Body text (undated page, fetched 2026-10-16)
Confidence: Verified verbatim
Squarespace's export leaves most of your site behind: only layout pages, one blog, text and image blocks move.
Squarespace Help Center, Export your site content · undated page, seen 2026 · the vendor’s own data · support.squarespace.com ↗
The sentence in the source
“Layout pages; One blog page, including its posts and up to 1,000 comments per post; Text and image blocks”
Where: "What exports" list (undated page, fetched 2026-10-16)
Confidence: Verified verbatim (list items)
Squarespace will not export store pages, portfolios, headers and footers, video or audio, drafts, styling or custom CSS.
Squarespace Help Center, Export your site content · undated page, seen 2026 · the vendor’s own data · support.squarespace.com ↗
The sentence in the source
“Most other page types, including album, cover, index, info, calendar, portfolio, and store pages; Content in page-specific headers, footers, and sidebars; More than one blog page; Dropdowns; Audio, video, and product blocks; Drafts; Style settings; Custom CSS”
Where: "What doesn't export" list (undated page, fetched 2026-10-16)
Confidence: Verified verbatim
Even builders that do let you download code strip out the dynamic parts: Webflow's export drops the CMS, and search and forms stop working.
Webflow Help, How do I export my Webflow site code? · undated page, seen 2026 · the vendor’s own data · help.webflow.com ↗
The sentence in the source
“CMS content and functionality (including Collection items and Collection lists) … site search and forms (including file upload and reCAPTCHA) will not work on exported sites”
Where: "What's not exported" section (undated page, fetched 2026-10-16)
Confidence: Verified verbatim; the plan-level statement is a minor paraphrase
Duda, a builder many agencies use, exports a zip only on Agency plans and above, and contact forms stop working after export.
Duda Support, Site Export · undated page, seen 2026 · the vendor’s own data · support.duda.co ↗
The sentence in the source
“Available on Agency plan and higher.”
Where: Top of article and "What is not exported" (undated page, fetched 2026-10-16)
Confidence: Plan quote verbatim; the exclusions are a minor paraphrase of the fetched list
About 4 in 10 websites run WordPress; Wix is on 4.2% of websites, Squarespace on 2.4%, and GoDaddy Website Builder on 0.6%.
W3Techs, Usage statistics of content management systems · 2026 · w3techs.com ↗
The sentence in the source
“WordPress is used by 39.7% of all the websites, that is a content management system market share of 58.4%.”
Where: Headline sentence and usage table, report dated 16 October 2026
Confidence: Verified verbatim (headline); table figures verified as listed
Roughly half of websites are built on some CMS or builder.
HTTP Archive, Web Almanac 2024, CMS chapter · 2024 · almanac.httparchive.org ↗
The sentence in the source
“As of June 2024, 51% of the websites in the Web Almanac's desktop dataset were powered by a CMS”
Where: "CMS Adoption" section, Figure 12.1 (published 2024-11-11; data June 2024)
Confidence: Verified verbatim
On mobile in 2024, only 40% of WordPress sites passed Google's Core Web Vitals, up from 28% the year before.
HTTP Archive, Web Almanac 2024, CMS chapter · 2024 · almanac.httparchive.org ↗
The sentence in the source
“WordPress had a 28% pass rate in 2023 and 40% in 2024”
Where: "Core Web Vitals" section, Figure 12.7 caption (data June 2024)
Confidence: The sheet marks this row partial: the WordPress figure is verified verbatim from the caption; the other builders' pass rates were read from a figure description, not prose, and are left out of this record
For comparison, about half of all websites passed Core Web Vitals in mid-2024.
HTTP Archive, Web Almanac 2024, CMS chapter · 2024 · almanac.httparchive.org ↗
The sentence in the source
“For reference, the global CWV pass rate for all origins as of June 2024 is 51%.”
Where: "Core Web Vitals" section
Confidence: Verified verbatim
A typical Squarespace page is the heaviest of the big platforms, about 3 MB on mobile; a typical WordPress page is about 2 MB.
HTTP Archive, Web Almanac 2024, CMS chapter · 2024 · almanac.httparchive.org ↗
The sentence in the source
“WordPress has a median of 2,252 KB on desktop and 2,047 KB on mobile; Squarespace continues to deliver the heaviest median page weight of ~3 MB compared to ~3.5MB in 2022.”
Where: "Resource weights" section, Figures 12.19–12.20
Confidence: WordPress and Squarespace quotes verbatim; the sheet's Wix, Joomla and Drupal figures (from a caption) are left out of this record
The median web page weighs about 2.3 MB on a phone; a lean page built by hand can be a fraction of that, and that is a measurement to make, not a claim to take.
HTTP Archive, Web Almanac 2024, Page Weight chapter · 2024 · almanac.httparchive.org ↗
The sentence in the source
“the median page weight for a desktop page was 2,652 KB … the median page weight for a mobile page was 2,311 KB”
Where: Figures 16.15 and 16.37
Confidence: Verified verbatim
Google says Core Web Vitals feed into its ranking systems and recommends passing them.
Google Search Central, Understanding Google Page Experience; Understanding Core Web Vitals and Google search results · 2025 · developers.google.com ↗
The sentence in the source
“Core Web Vitals are used by our ranking systems. … We highly recommend site owners achieve good Core Web Vitals for success with Search.”
Where: Intro and FAQ of each document (last updated 2025-12-10); the second document is https://developers.google.com/search/docs/appearance/core-web-vitals
Confidence: Verified verbatim
Google also says that passing Core Web Vitals does not guarantee a top ranking.
Google Search Central, Understanding Google Page Experience · 2025 · developers.google.com ↗
The sentence in the source
“There is no single signal. Our core ranking systems look at a variety of signals that align with overall page experience. … doesn't guarantee that your pages will rank at the top of Google Search results.”
Where: FAQ section (last updated 2025-12-10)
Confidence: Verified verbatim
Email on your own domain, spoofing, fraud losses
Americans reported over $3 billion in losses to fake business emails (business email compromise) to the FBI for 2025.
FBI Internet Crime Complaint Center (IC3), 2025 Internet Crime Report · published 2026 · data 2025 · ic3.gov ↗
The sentence in the source
“Business Email Compromise $3,046,598,558”
Where: p. 8, losses by crime type; the three-year table on p. 26
Confidence: Verified verbatim
Nearly 25,000 businesses and individuals filed FBI complaints about email-impersonation fraud in 2025.
FBI Internet Crime Complaint Center (IC3), 2025 Internet Crime Report · published 2026 · data 2025 · ic3.gov ↗
The sentence in the source
“Business Email Compromise 24,768”
Where: p. 7, complaints by crime type; p. 25 for 2024
Confidence: Verified verbatim
Phishing and spoofed emails drew over 191,000 complaints to the FBI in 2025.
FBI Internet Crime Complaint Center (IC3), 2025 Internet Crime Report · published 2026 · data 2025 · ic3.gov ↗
The sentence in the source
“Phishing/Spoofing 191,561 … Phishing/Spoofing $215,843,126”
Where: pp. 7–8; the three-year tables on pp. 25–26
Confidence: Verified verbatim
Total reported cybercrime losses in the US passed $20 billion in 2025, up 26% in one year.
FBI Internet Crime Complaint Center (IC3), 2025 Internet Crime Report · published 2026 · data 2025 · ic3.gov ↗
The sentence in the source
“1,008,597 complaints; $20.877 billion in losses; 26% increase in losses from 2024”
Where: p. 6 (and the chart labels on p. 4)
Confidence: Verified verbatim
Roughly 7 out of 10 fake-boss and fake-vendor scam emails are sent from free webmail accounts like Gmail, so a business that itself emails from Gmail looks exactly like the scammers do.
APWG (Anti-Phishing Working Group), Phishing Activity Trends Report, 4th Quarter 2025 · published 2026 · data 2025 · docs.apwg.org ↗
The sentence in the source
“Fortra observed that 69 percent of BEC attacks in Q4 2025 were launched using a free webmail domain … The remaining 31 percent of BEC attacks in Q4 2025 utilized nonwebmail domains.”
Where: p. 13 (Fortra data)
Confidence: Verified verbatim; APWG is an industry association reporting its members' (Fortra's) observed attacks — not a vendor's own data about its product
When a scammer impersonates a business by email and asks for a wire transfer, the average ask is about $50,000.
APWG (Anti-Phishing Working Group), Phishing Activity Trends Report, 4th Quarter 2025 · published 2026 · data 2025 · docs.apwg.org ↗
The sentence in the source
“the average amount requested in wire transfer BEC attacks in Q4 2025 was $50,297”
Where: p. 11
Confidence: Verified verbatim; APWG is an industry association reporting its members' (Fortra's) observed attacks — not a vendor's own data about its product
Phishing attacks ran at about 3.8 million for the year 2025.
APWG (Anti-Phishing Working Group), Phishing Activity Trends Report, 4th Quarter 2025 · published 2026 · data 2025 · docs.apwg.org ↗
The sentence in the source
“During 2025, APWG observed 3.8 million phishing attacks, which was up slightly from 3.76 million in 2024. … In the fourth quarter of 2025, APWG observed 853,244 phishing attacks.”
Where: p. 3
Confidence: Verified verbatim; APWG is an industry association reporting its members' (Fortra's) observed attacks — not a vendor's own data about its product
People, not software flaws, were involved in about 6 of every 10 confirmed data breaches.
Verizon, 2026 Data Breach Investigations Report, executive summary · published 2026 · data 2024–2025 · verizon.com ↗
The sentence in the source
“Human element was present in 62% of breaches, a slight increase from the previous year's 60%.”
Where: Key findings; the dataset covers October 2024 through November 2025
Confidence: Verified verbatim
Phishing was the entry point in about 1 in 6 breaches, and tricking people by email is one of the top three ways companies get breached.
Verizon, 2026 Data Breach Investigations Report, executive summary · published 2026 · data 2024–2025 · verizon.com ↗
The sentence in the source
“Phishing remained at 16% like the previous year. … Social Engineering was our third most common breach pattern, representing 16% of all breaches.”
Where: Key findings / patterns section
Confidence: Verified verbatim (the first quote is clipped from a longer sentence about pretexting at 6%)
In the worst cases, a single breach cost a small business more than 7% of its annual revenue.
Verizon, 2026 DBIR Breach Impact Study (infographic, about 70,000 cyber-insurance claims) · 2026 · verizon.com ↗
The sentence in the source
“In extreme cases (top 2.5%), financial loss due to a data breach accounted for more than 7% of SMBs' revenue.”
Where: Single-page infographic
Confidence: Verified verbatim; the infographic states no data period
Email fraud (business email compromise) is a small slice of the threats Microsoft sees, but it ends up being the actual outcome of an attack more often than ransomware does.
Microsoft, Digital Defense Report 2025 · published 2025 · data 2024–2025 · the vendor’s own data · cdn-dynmedia-1.microsoft.com ↗
The sentence in the source
“While business email compromise (BEC) represented just 2% of total threats observed over the past year, its impact is disproportionately high... … In fact, BEC was a more frequent outcome in attacks (21%) than ransomware (16%)”
Where: p. 38 (data July 2024 – June 2025)
Confidence: Verified verbatim
Over a quarter of the breaches Microsoft investigated started with a phishing email or someone being tricked.
Microsoft, Digital Defense Report 2025 · published 2025 · data 2024–2025 · the vendor’s own data · cdn-dynmedia-1.microsoft.com ↗
The sentence in the source
“28% of breaches were initiated through phishing or social engineering”
Where: p. 9
Confidence: Verified verbatim
Nearly half the organizations Microsoft's incident responders helped were small-to-mid-sized businesses with revenue under $50 million.
Microsoft, Digital Defense Report 2025 · published 2025 · data 2024–2025 · the vendor’s own data · cdn-dynmedia-1.microsoft.com ↗
The sentence in the source
“Almost half (48%) of organizations whose size is known had an annual revenue of USD 50 million or less.”
Where: p. 28
Confidence: Verified verbatim
Among UK businesses that were attacked, phishing was by far the most common attack and the one most often named as the most disruptive; about 1 in 8 businesses had someone impersonate them by email or online.
UK Department for Science, Innovation and Technology / Home Office, Cyber Security Breaches Survey 2025/2026 · published 2026 · data 2025 · gov.uk ↗
The sentence in the source
“Phishing attacks remained the most prevalent type of breach or attack by far (experienced by 38% of businesses and 25% of charities) … continued to be ascribed as the most disruptive type of breach or attack (69% of businesses and charities that experienced a breach or attack) … Impersonation breaches or attacks have decreased to 12% this year, down from 17% in 2023.”
Where: Summary, "Identification of cyber security breaches and attacks" (fieldwork August–December 2025)
Confidence: Verified verbatim
About 43% of UK businesses, roughly 612,000 of them, found at least one cyber attack in the past year; micro businesses were hit at 42% and small ones at 46%.
UK Department for Science, Innovation and Technology / Home Office, Cyber Security Breaches Survey 2025/2026 · published 2026 · data 2025 · gov.uk ↗
The sentence in the source
“Just over four in ten businesses (43%)... equates to approximately 612,000 UK (United Kingdom) businesses … micro (42%) and small (46%) businesses”
Where: Summary
Confidence: Verified with minor paraphrase (stitched from two adjacent sentences); UK data
Only about 86% of the email Cloudflare processed passed DMARC checks and 4% outright failed, so a meaningful share of mail arriving in inboxes cannot be tied to the domain it claims to come from.
Cloudflare blog, Launching email security insights on Cloudflare Radar · 2024 · the vendor’s own data · blog.cloudflare.com ↗
The sentence in the source
“over 93% of messages passed SPF authentication, while just 2.7% failed. … An average of 88.5% of messages passed DKIM evaluation in February, while just 2.1% failed. … For DMARC, 86.5% of messages passed authentication, while 4.2% failed”
Where: SPF/DKIM/DMARC section (published 2024-03-08; data February 2024)
Confidence: Verified verbatim; message-level, not domain-level, and 2024 data
Gmail says it blocks nearly 15 billion unwanted emails a day and cut the unauthenticated mail reaching its users by 75% by requiring senders to authenticate.
Google, The Keyword blog, New Gmail protections for a safer, less spammy inbox · 2023 · the vendor’s own data · blog.google ↗
The sentence in the source
“block nearly 15 billion unwanted emails … stop more than 99.9% of spam, phishing and malware … the number of unauthenticated messages Gmail users receive plummet by 75%”
Where: Body text (published 2023-10-03)
Confidence: Verified verbatim; older (2023)
Compromised sites, outdated software, headers
Of the hacked websites Sucuri cleaned up in 2023, roughly four in ten were running out-of-date software when they got infected.
Sucuri, 2023 Hacked Website & Malware Threat Report · published 2024 · data 2023 · the vendor’s own data · sucuri.net ↗
The sentence in the source
“In 2023, 39.1% of all Content Management System (CMS) applications were outdated at the time of infection.”
Where: Section "Outdated CMS Detection"
Confidence: Verified verbatim; Sucuri's own cleanup data
About half of hacked sites had a hidden backdoor left in them, so the attacker can walk back in after a surface cleanup.
Sucuri, 2023 Hacked Website & Malware Threat Report · published 2024 · data 2023 · the vendor’s own data · sucuri.net ↗
The sentence in the source
“In 2023, 49.21% of compromised websites were discovered to contain at least one backdoor at the time of infection.”
Where: Section "Backdoors"
Confidence: Verified verbatim; Sucuri's own cleanup data
One in five hacked sites was being used to push spam links, which quietly wrecks its Google ranking.
Sucuri, 2023 Hacked Website & Malware Threat Report · published 2024 · data 2023 · the vendor’s own data · sucuri.net ↗
The sentence in the source
“20.30% of all remediated sites were affected by some form of SEO spam.”
Where: Section "SEO Spam"
Confidence: Verified verbatim; Sucuri's own cleanup data
Over 11,000 new security holes were found in the WordPress ecosystem in 2025, up 42% from the year before.
Patchstack, State of WordPress Security in 2026 · published 2026 · data 2025 · the vendor’s own data · patchstack.com ↗
The sentence in the source
“Overall 11,334 new vulnerabilities were found in the WordPress ecosystem in 2025”
Where: Whitepaper body, vulnerability totals section
Confidence: Verified verbatim (the +42% is the page's own framing, confirmed in the fetch); Patchstack's own database
Nearly all of those holes were in plugins, not in WordPress itself.
Patchstack, State of WordPress Security in 2026 · published 2026 · data 2025 · the vendor’s own data · patchstack.com ↗
The sentence in the source
“91% of new vulnerabilities were found in plugins, and 9% were found in themes. … There were only 6 vulnerabilities reported in the WordPress core, but these were low priority issues.”
Where: Whitepaper body, vulnerability totals section
Confidence: Verified verbatim; Patchstack's own database
Almost half the time, the plugin maker had not released a fix by the day the hole was made public.
Patchstack, State of WordPress Security in 2026 · published 2026 · data 2025 · the vendor’s own data · patchstack.com ↗
The sentence in the source
“46% of vulnerabilities did not receive a fix from the developer in time for public disclosure.”
Where: Whitepaper body, vulnerability totals section
Confidence: Verified verbatim; Patchstack's own database
When a serious plugin hole is disclosed, attackers typically start using it within hours, not weeks.
Patchstack, State of WordPress Security in 2026 · published 2026 · data 2025 · the vendor’s own data · patchstack.com ↗
The sentence in the source
“approximately half of high impact vulnerabilities get exploited within 24 hours … the weighted median time to first exploit is 5 hours.”
Where: Exploitation timing section
Confidence: Verified verbatim; Patchstack's own telemetry
The year before, 2024, saw about 8,000 new WordPress vulnerabilities, 96% of them in plugins, and 43% needed no login to exploit.
Patchstack, State of WordPress Security in 2025 · published 2025 · data 2024 · the vendor’s own data · patchstack.com ↗
The sentence in the source
“In 2024, 7,966 new vulnerabilities were found in the WordPress ecosystem, primarily in third-party plugins. … 96% of the vulnerabilities were uncovered in plugins, and 4% were found in themes. … 43% of new vulnerabilities found in 2024 did not require any authentication to be exploited.”
Where: Whitepaper body
Confidence: Verified verbatim; Patchstack's own database
One security vendor alone blocked 48 billion attempts to exploit WordPress sites and 55 billion password-guessing attempts in a single year.
Wordfence, 2024 Annual WordPress Security Report · published 2025 · data 2024 · the vendor’s own data · wordfence.com ↗
The sentence in the source
“Out of those, Wordfence blocked over 48 billion requests targeting WordPress vulnerabilities, … Wordfence also blocked over 55 billion password-hacking attempts,”
Where: pp. 48 and 50
Confidence: Verified verbatim; Wordfence's own customer telemetry, not web-wide
Small businesses showed up as breach victims almost four times as often as large ones in Verizon's 2025 data.
Verizon, 2025 Data Breach Investigations Report, SMB Snapshot · published 2025 · data 2023–2024 · verizon.com ↗
The sentence in the source
“3,049 incidents, 2,842 with confirmed data disclosure … there are almost four times the number of SMB victims than there are large organizations.”
Where: Table 1 and the accompanying text (data November 2023 – October 2024)
Confidence: Verified verbatim
When a small business is breached, ransomware is involved almost nine times out of ten.
Verizon, 2025 Data Breach Investigations Report, SMB Snapshot · published 2025 · data 2023–2024 · verizon.com ↗
The sentence in the source
“SMBs experienced Ransomware-related breaches to the tune of 88% overall.”
Where: Body text
Confidence: Verified verbatim
Small-business breaches are almost always outsiders, almost always after money, and most often start with a stolen password.
Verizon, 2025 Data Breach Investigations Report, SMB Snapshot · published 2025 · data 2023–2024 · verizon.com ↗
The sentence in the source
“External (98%), Internal (2%), Partner (1%) … Financial (99%) … the primary hacking variety for both is the Use of stolen credentials, at 32% in large organizations and 33% in SMBs.”
Where: Table 1 (SMB row) and text
Confidence: Verified verbatim
Across all organizations, breaking in through an unpatched software hole was the way in for one breach in five, up a third from the prior year.
Verizon, 2025 Data Breach Investigations Report, press release · published 2025 · data 2023–2024 · verizon.com ↗
The sentence in the source
“exploitation of vulnerabilities (20%) … This initial attack vector saw a 34% increase”
Where: Press release body (data November 2023 – October 2024)
Confidence: Verified verbatim (press release, not the full PDF)
Americans reported $16.6 billion in internet-crime losses to the FBI in 2024, a third more than the year before.
FBI Internet Crime Complaint Center (IC3), 2024 Internet Crime Report · published 2025 · data 2024 · ic3.gov ↗
The sentence in the source
“In 2024, complaints totaled 859,532 … losses of $16.6 billion … representing a 33 percent increase from 2023”
Where: p. 4; BEC $2,770,151,146 on p. 11
Confidence: Verified verbatim
Half of UK small businesses (10–49 staff) and four in ten micro businesses were hit by a cyber breach or attack in the year to late 2024.
UK Department for Science, Innovation and Technology, Cyber Security Breaches Survey 2025 · published 2025 · data 2024 · gov.uk ↗
The sentence in the source
“micro businesses now 41%, down from 47% in 2024 and small businesses now 50% down from 58% in 2024 … Just over four in ten businesses (43%)”
Where: Chapter 4, Figure 4.1 discussion; Summary (fieldwork late 2024)
Confidence: Verified verbatim (UK data, not US)
The average UK business that had a damaging breach put the cost at about £3,550 once the zero-cost cases are excluded.
UK Department for Science, Innovation and Technology, Cyber Security Breaches Survey 2025 · published 2025 · data 2024 · gov.uk ↗
The sentence in the source
“the average cost of the most disruptive breach for each business in the last 12 months was £1,600 for businesses … the average cost of the most disruptive breach was £3,550 for businesses”
Where: Summary, identification of breaches
Confidence: Verified verbatim; UK data
Nearly every website now loads over HTTPS, so a site without the padlock stands out.
HTTP Archive, Web Almanac 2025, Security chapter · published 2026 · data 2025 · almanac.httparchive.org ↗
The sentence in the source
“This number rises from 95.6% to 97.3% on mobile.”
Where: Section "Transport security" (published January 2026; crawl 2025)
Confidence: Verified verbatim
Only about one site in five sets a Content-Security-Policy header, the main defense against injected scripts and clickjacking.
HTTP Archive, Web Almanac 2025, Security chapter · published 2026 · data 2025 · almanac.httparchive.org ↗
The sentence in the source
“The adoption of CSP continued increasing from 18.5% last year to 21.9% this year, an increase of close to 20%.”
Where: Section "Content Security Policy"
Confidence: Verified verbatim
Only about a third of sites set the headers that stop their pages being framed by attackers or that force encrypted connections.
HTTP Archive, Web Almanac 2025, Security chapter · published 2026 · data 2025 · almanac.httparchive.org ↗
The sentence in the source
“showing steady increases to approximately 35 percent of sites. … up to 36% of all pages visited on mobile.”
Where: Figure 9.27 description; HSTS section
Confidence: Verified with minor paraphrase (the Figure 9.27 caption groups X-Frame-Options and HSTS at about 35%)
Fewer than half of sites set even the simplest hardening header.
HTTP Archive, Web Almanac 2025, Security chapter · published 2026 · data 2025 · almanac.httparchive.org ↗
The sentence in the source
“with `X-Content-Type-Options` leading the group at nearly 50 percent adoption by 2025.”
Where: Figure 9.27 description
Confidence: Verified verbatim
Listings, reviews, wrong hours, phones, speed
Three out of four Americans read online reviews often or every time before choosing a local business.
BrightLocal, Local Consumer Review Survey 2024 · published 2024 · data likely 2023 · the vendor’s own data · brightlocal.com ↗
The sentence in the source
“the percentage of consumers 'always' or 'regularly' reading online reviews has held fast (75% in 2024 against 76% in 2023)”
Where: "Reading reviews" section (published 2024-03-06; panel of 1,141 US consumers; fieldwork likely late 2023)
Confidence: Verified verbatim; a vendor survey with its sample and method published
Only 4 in 100 people never look at reviews at all.
BrightLocal, Local Consumer Review Survey 2025 · published 2025 · data likely 2024 · the vendor’s own data · brightlocal.com ↗
The sentence in the source
“just 4% of consumers say they 'never' read online business reviews.”
Where: Review-reading frequency (published 2025-01-29; 1,026 US adults; fieldwork likely late 2024)
Confidence: Verified verbatim; a vendor survey with its sample and method published
Nearly 9 in 10 people will use a business that answers all its reviews; fewer than half will use one that ignores them.
BrightLocal, Local Consumer Review Survey 2024 · published 2024 · data likely 2023 · the vendor’s own data · brightlocal.com ↗
The sentence in the source
“88% of consumers would use a business that replies to all of its reviews … compared to just 47% who would use a business that doesn't respond to reviews at all”
Where: Review-response section
Confidence: Verified verbatim; a vendor survey with its sample and method published
Seven in ten people won't consider a business whose average rating is under three stars.
BrightLocal, Local Consumer Review Survey 2024 · published 2024 · data likely 2023 · the vendor’s own data · brightlocal.com ↗
The sentence in the source
“71% of consumers would not consider using a business with an average rating below three stars”
Where: Star-rating section
Confidence: Verified verbatim; a vendor survey with its sample and method published
One in five customers wants to see a review posted within the last two weeks before it counts.
BrightLocal, Local Consumer Review Survey 2025 · published 2025 · data likely 2024 · the vendor’s own data · brightlocal.com ↗
The sentence in the source
“20% of consumers feel that reviews as recent as two weeks old are impactful to their decisions.”
Where: Review recency
Confidence: Verified verbatim; a vendor survey with its sample and method published
Half of people trusted online reviews as much as a friend's recommendation in 2024; 42% did in 2025, down from 79% in 2020.
BrightLocal, Local Consumer Review Survey 2024 and 2025 · published 2025 · data likely 2023–2024 · the vendor’s own data · brightlocal.com ↗
The sentence in the source
“50% of consumers trust reviews as much as personal recommendations from friends and family. … 79% … this now sits at 42% in 2025.”
Where: Trust section of each edition; the 2024 edition is https://www.brightlocal.com/research/local-consumer-review-survey-2024/
Confidence: Verified verbatim; cite with the year, the drop is sharp; a vendor survey with its sample and method published
Six in ten people would avoid a business after finding wrong information about it online.
BrightLocal, Local Business Discovery & Trust Report 2023 · 2023 · the vendor’s own data · brightlocal.com ↗
The sentence in the source
“62% of consumers would avoid using a business if they found incorrect information online.”
Where: Reactions to incorrect information (survey September 2023, 1,138 US consumers)
Confidence: Minor paraphrase; a vendor survey with its sample and method published
85% of people say a business's contact information and opening hours matter when they research it.
BrightLocal, Consumer Search Behavior · 2025 · the vendor’s own data · brightlocal.com ↗
The sentence in the source
“85% of consumers consider the presence of contact information and opening hours an important factor…”
Where: Decision factors (April 2025, 1,000 US adults)
Confidence: Verified verbatim; a vendor survey with its sample and method published
Three-quarters of people who search for something nearby on their phone walk into a business within a day, and more than a quarter buy.
Google, Think with Google, How Mobile Search Connects Consumers to Stores · 2016 · the vendor’s own data · thinkwithgoogle.com ↗
The sentence in the source
“of people who search on their smartphones for something nearby visit a business within a day … of those searches … result in a purchase.”
Where: Stat callouts, footnote 7 (data May 2016)
Confidence: Verified verbatim; data 2016; Google's own search data
Roughly a third of all phone searches are about a place.
Google, Think with Google, How Mobile Search Connects Consumers to Stores · 2016 · the vendor’s own data · thinkwithgoogle.com ↗
The sentence in the source
“30% of all mobile searches are related to location.”
Where: Footnote 4 (data April 2016); also https://www.thinkwithgoogle.com/_qs/documents/2272/1057-Insights-Right-Here-Download.pdf, footnote 1
Confidence: Verified verbatim; data 2016; Google's own search data
More than half of mobile visitors give up on a page that takes over three seconds to load.
Google, Think with Google, New Industry Benchmarks for Mobile Page Speed · published 2017 · data 2016 · the vendor’s own data · thinkwithgoogle.com ↗
The sentence in the source
“Yet 53% of mobile site visitors leave a page that takes longer than three seconds to load.”
Where: Intro, footnote 2 (published February 2017; Google internal data, March 2016)
Confidence: Verified verbatim; Google's own data, 2016
Going from a one-second load to seven seconds more than doubles the odds that a phone visitor leaves.
Google, Think with Google, New Industry Benchmarks for Mobile Page Speed · 2017 · the vendor’s own data · thinkwithgoogle.com ↗
The sentence in the source
“as page load time goes from one second to seven seconds, the probability of a mobile site visitor bouncing increases 113%”
Where: Body, footnote 6 (Google/SOASTA data 2017)
Confidence: Verified verbatim; Google's own data
Shaving a tenth of a second off mobile load lifted retail conversions 8.4% and order size 9.2% in a study of 37 sites.
Deloitte Digital with Google, Milliseconds Make Millions · published 2020 · data 2019 · deloitte.com ↗
The sentence in the source
“Based on a 0.1s natural mobile site speed improvement… … Retail conversions increased by 8.4% and average order value increased by 9.2%.”
Where: Key findings (37 sites, 30M+ sessions, late 2019); also https://web.dev/case-studies/milliseconds-make-millions
Confidence: Verified verbatim (HTML; the PDF fetch failed)
When people judged whether a website was trustworthy, nearly half of their comments were about how it looked.
Fogg et al., Stanford Persuasive Technology Lab with Consumer WebWatch, How Do People Evaluate a Web Site's Credibility? · 2002 · advocacy.consumerreports.org ↗
The sentence in the source
“…2,684 people evaluated the credibility of two live Web sites… … design look … present in 46.1 percent of the comments.”
Where: Summary page (published 2002-10-29)
Confidence: Verified verbatim; data 2002, a landmark study; it is about design quality in general, not templates
Google says customers are nearly 3 times more likely to see a business as reputable, 70% more likely to visit, and 50% more likely to consider buying when its Google Business Profile is complete.
Google Business Profile Help, Complete your Business Profile on Google · undated page, seen 2026 · the vendor’s own data · support.google.com ↗
The sentence in the source
“Customers are 2.7 times more likely to consider a business reputable if they find a complete Business Profile… … 70% more likely to visit and 50% more likely to consider purchasing…”
Where: Top of page (undated, current 2026)
Confidence: Verbatim, but no study or date is given — Google's own statement
The typical local business got about 1,260 Google profile views a month and 33 website clicks from its listing, in a study of 45,264 listings.
BrightLocal, Google My Business Insights Study · published 2019 · data 2017–2018 · the vendor’s own data · brightlocal.com ↗
The sentence in the source
“The average local business receives 1,260 views each month … 33 clicks each month from their GMB listing”
Where: Key findings (July 2019; data September 2017 – December 2018, 45,264 listings)
Confidence: Views and clicks verbatim; the sheet's direction-request figure (partial) is left out; a dated benchmark; a vendor's own data
Over 8 in 10 Americans read reviews before buying something for the first time, and 4 in 10 almost always do.
Pew Research Center, Online Shopping and E-Commerce · published 2016 · data 2015 · pewresearch.org ↗
The sentence in the source
“Fully 82% of U.S. adults say they at least sometimes read online customer ratings or reviews before purchasing items for the first time, including 40% who say they always or almost always do so.”
Where: Chapter "Online reviews and ratings" (survey November–December 2015, 4,787 adults)
Confidence: Verified verbatim; data 2015, general purchases
Nine in ten Americans carry a smartphone, and 16% rely on it as their only way online.
Pew Research Center, Mobile Fact Sheet · 2025 · pewresearch.org ↗
The sentence in the source
“About nine-in-ten (91%) own a smartphone … 16% of U.S. adults are 'smartphone-only' internet users.”
Where: Top of the fact sheet (November 2025; survey February–June 2025)
Confidence: Verified verbatim
One extra star on Yelp meant roughly 5–9% more revenue for an independent restaurant.
Michael Luca, Harvard Business School, Reviews, Reputation, and Revenue: The Case of Yelp.com · published 2011 · data 2003–2009 · papers.ssrn.com ↗
The sentence in the source
“a one-star increase in Yelp rating leads to a 5-9 percent increase in revenue”
Where: Abstract (SSRN 2011, revised 2016; Seattle restaurants 2003–2009)
Confidence: Verified verbatim; independent restaurants only
Real photos, bios, plain copy, accessibility, downtime
Visitors look hard at photos of your real people and real products, and skip generic stock imagery.
Jakob Nielsen, Nielsen Norman Group, Photos as Web Content · published 2010 · data 2006–2010 · nngroup.com ↗
The sentence in the source
“Photos of products and real people (as opposed to stock photos of models) … big feel-good images that are purely decorative … jazzed-up = ignored”
Where: Opening section and the "Images That Are Ignored / Images That Are Looked At" examples (published 2010-10-31; eyetracking data ~2006–2010; page last reviewed 2026-08-13)
Confidence: Verified verbatim for the quoted phrases; the popular "big fluffy filler" wording does not appear on the page
Staff headshots on a team page pulled more eyeball time than the bios next to them, even though the bios were far larger.
Jakob Nielsen, Nielsen Norman Group, Photos as Web Content · published 2010 · data 2006–2010 · nngroup.com ↗
The sentence in the source
“Portrait photos of FreshBooks staff drew 10% more viewing time than the bios, even though the bios took up 316% more space”
Where: FreshBooks team-page example, mid-article
Confidence: Minor paraphrase (the fetch restated the numbers; 10% and 316% confirmed on the page)
On product pages, people spend most of their time reading the words, not the pictures, so the copy still has to carry the load.
Jakob Nielsen, Nielsen Norman Group, Photos as Web Content · published 2010 · data 2006–2010 · nngroup.com ↗
The sentence in the source
“photos received 18% of viewing time and text received 82%”
Where: Amazon product-page example
Confidence: Minor paraphrase (numbers confirmed)
Visitors to your About page want to know what you do, where you are and how to reach you, and they check you against outside reviews before trusting your own claims.
Anna Kaley and Jakob Nielsen, Nielsen Norman Group, 'About Us' Information on Websites · published 2019 · data 2018–2019 · nngroup.com ↗
The sentence in the source
“At this point, I'd just Google it. … the word reviews was used over 40 times”
Where: Methodology paragraph; "Users Are More Skeptical" section; the guidelines list (published 2019-05-26)
Confidence: Verified verbatim for the quotes; the counts are a minor paraphrase
Almost every website home page fails basic accessibility checks, averaging dozens of errors each.
WebAIM, The WebAIM Million, 2026 report · 2026 · webaim.org ↗
The sentence in the source
“95.9% of home pages had detected WCAG 2 failures ... 56.1 per page (56,114,377 total), up 10.1% from 51 per page in 2025”
Where: "Detected Errors" and "WCAG Conformance" sections (8th annual report; data February 2026; page updated 2026-03-30)
Confidence: Verified verbatim
Six simple mistakes account for almost all accessibility failures, and the biggest is text that is hard to read against its background.
WebAIM, The WebAIM Million, 2026 report · 2026 · webaim.org ↗
The sentence in the source
“96% of all errors detected fall into these six categories”
Where: "WCAG Conformance" table of the most common failure types; "Low Contrast Text" section
Confidence: Verified verbatim (the 96% quote); the per-category percentages confirmed from the page table
More than one in four American adults lives with a disability, so an inaccessible site turns away a large slice of any local market.
CDC press release, CDC Data Shows Over 70 Million U.S. Adults Reported Having a Disability · published 2024 · data 2022 · cdc.gov ↗
The sentence in the source
“more than 1 in 4 ... over 70 million”
Where: Lead paragraph (published 2024-07-16; data from the 2022 BRFSS)
Confidence: Verified verbatim; the oft-quoted 28.7% is not on this page
The federal government now requires state and local government websites to meet WCAG 2.1 AA, with deadlines in 2027 and 2028; the rule does not directly cover private businesses.
US Department of Justice, ada.gov, Fact Sheet: New Rule on the Accessibility of Web Content and Mobile Apps Provided by State and Local Governments · 2024 · ada.gov ↗
The sentence in the source
“Web Content Accessibility Guidelines (WCAG) Version 2.1, Level AA is the technical standard ... within three or four years of when the rule was published on April 24, 2024”
Where: "What is the technical standard" and "When do governments have to comply" sections (page dated 2024-04-08; final rule published 2024-04-24; the deadlines reflect an interim final rule of 2026-04-20)
Confidence: Verified verbatim; the page addresses private companies only when they build content on a government's behalf
Website accessibility lawsuits are a recurring thing: over 2,400 were filed in federal court in 2024 alone, and that excludes state courts.
Seyfarth Shaw, ADA Title III blog (Launey and Vu), Federal Court Website Accessibility Lawsuit Filings Continue to Decrease in 2024 · published 2025 · data 2024 · adatitleiii.com ↗
The sentence in the source
“are not included in our federal lawsuit numbers”
Where: Opening paragraph and the bar graph (April 2025)
Confidence: Verified (counts on the page); the 2023 figure is internally inconsistent on the page (2,794 in the graph, 2,749 in the text); the publication date is inferred from the URL
Nearly three in ten working-age American adults read at the lowest literacy level, so plain, short copy is not dumbing down; it is reaching customers.
National Center for Education Statistics (NCES/IES) press release, U.S. Adults Score on Par With International Average in Literacy Skills, Below International Average in Numeracy · published 2024 · data 2023 · ies.ed.gov ↗
The sentence in the source
“adults scoring Level 1 or below”
Where: Literacy results paragraph (published 2024-12-10; data from the 2023 PIAAC)
Confidence: Verified (figures on the page); the companion NCES report page does not display the percentages itself
Stanford's credibility guidelines say to show real people, give a real address, keep content visibly current, and fix every typo and broken link.
Stanford Persuasive Technology Lab, Stanford Guidelines for Web Credibility · published 2002 · data 1999–2002 · credibility.stanford.edu ↗
The sentence in the source
“Show that honest and trustworthy people stand behind your site. … Update your site's content often (at least show it's been reviewed recently). … Avoid errors of all types, no matter how small they seem.”
Where: Guidelines 2, 4, 5, 8 and 10 (updated June 2002; research 1999–2002)
Confidence: Verified verbatim
Among large data-center operators surveyed, more than half said their last serious outage cost over $100,000 and one in five said over $1 million; that is enterprise data, not small-business data, and no small-business downtime-cost study was found.
Uptime Institute Intelligence, Annual outage analysis 2025 (Keynote Report 173) · published 2025 · data 2024 · intelligence.uptimeinstitute.com ↗
The sentence in the source
“overall outage frequency and the general level of reported severity continue to decline ... Power issues remain the most common cause of serious and severe data center outages.”
Where: Public summary and key findings (published 2025-05-02; the full report is paywalled)
Confidence: Minor paraphrase for the cost figures (seen in the summary); the quotes verbatim
Rule for every line above: the primary document was opened and the figure was seen in it. Old landmark studies are in because they are the primaries everyone else quotes, and they are dated as such. Back to what we do.