Your website is not one thing.

It’s a pile of decisions: CMS, theme, plugins, hosting, analytics, pixels, forms, fonts, scripts, embeds, cookie tools, and whatever somebody added during a campaign two years ago and forgot to remove.

That pile matters. It affects speed, search visibility, privacy risk, maintenance cost, and whether a normal person can use the site without waiting, tapping twice, or giving up.

This data roundup is built for web professionals, business owners, and anyone who has ever opened a tag manager container and wondered why there are 47 things firing on one contact page.

CMS statistics: what businesses are actually building on

A CMS used to mean “the thing marketing logs into.” Now it’s the operating layer for publishing, SEO, media, page templates, forms, structured data, user roles, ecommerce, redirects, and often half the performance problems.

1. CMS-driven sites account for over 54% of observed websites. HTTP Archive’s 2025 CMS chapter says CMSs are now used by the majority of sites, making them default web infrastructure rather than a niche publishing choice. (HTTP Archive)

2. W3Techs reports that 30% of websites use none of the CMSs it monitors. That doesn’t mean those sites have no backend at all. It means the monitored CMS fingerprint wasn’t detected, which may include custom builds, static sites, internal platforms, or systems outside its tracking list. (W3Techs)

3. WordPress is used by 41.5% of all websites. W3Techs gives WordPress a 59.3% CMS market share, which explains why plugin quality, WordPress hosting, and theme decisions still shape a huge slice of the web. (W3Techs)

4. HTTP Archive found WordPress powers roughly 64% of CMS-driven sites. Its 2025 analysis also says WordPress growth has slowed to under one percentage point year over year, more like saturation than collapse. (HTTP Archive)

5. Shopify is the second-largest CMS in W3Techs’ June 2026 data at 5.2% of all websites and 7.5% CMS share. That number matters because ecommerce platforms are not only storefronts. They’re checkout, analytics, theme code, app ecosystems, product feeds, email capture, and conversion testing all in one stack. (W3Techs)

6. Wix is used by 4.3% of all websites, with a 6.1% CMS market share. HTTP Archive describes hosted CMS platforms as more concentrated in North America and Western Europe, which matches what many local service businesses actually buy. (W3Techs, HTTP Archive)

7. Squarespace is used by 2.5% of all websites and has 3.5% CMS share. For small businesses, that is big enough to treat Squarespace migrations, SEO audits, and performance checks as normal work, not edge cases. (W3Techs)

8. Drupal is only 0.7% of all websites in W3Techs data, but HTTP Archive says it is disproportionately represented among high-traffic sites. HTTP Archive reports Drupal is about 6% to 7% of CMS usage among the top 10,000 websites, far above its overall footprint. (W3Techs, HTTP Archive)

The lesson is not “use WordPress” or “avoid Wix.” The real lesson is that CMS choice affects who can update the site, how fast work gets done, how expensive changes become, and how much technical cleanup you’ll need later.

JavaScript statistics: old libraries, new frameworks, same payload problem

JavaScript is useful. It also has a cost that is easy to hide in a proposal and hard to ignore on a five-year-old phone.

9. W3Techs reports that 22.1% of websites use none of the JavaScript libraries it monitors. The other side of that number is more useful: most websites depend on at least one detectable library. (W3Techs)

10. jQuery is used by 68.1% of all websites. It also holds 87.3% market share among sites using a monitored JavaScript library, which means a lot of the web still runs on older, battle-tested frontend plumbing. (W3Techs)

11. Bootstrap appears on 15.6% of all websites. That makes it the second-largest monitored JavaScript library in W3Techs’ June 2026 dataset. (W3Techs)

12. React is used by 6.1% of all websites. That number is smaller than the developer conversation can make it feel, which is a good reminder that GitHub trends and small business websites are not the same market. (W3Techs)

13. Next.js is used by 2.8% of all websites. That is real adoption, but still far behind jQuery, Bootstrap, and even Underscore in W3Techs’ monitored library list. (W3Techs)

14. HTTP Archive found the median desktop page made 23 JavaScript requests, while the median mobile page made 22. Request count is not the only performance issue, but it is a practical clue when auditing a bloated page. (HTTP Archive)

15. The median mobile home page grew from 845 KB in July 2015 to 2,362 KB in July 2025. HTTP Archive calls that a 202.8% increase over the decade. (HTTP Archive)

16. The median desktop home page saw a 110.2% size increase over the same period. That is not just prettier imagery. It’s more JavaScript, more media, more fonts, more scripts, and more decisions stacked together. (HTTP Archive)

17. The median home page size grew 7.8% year over year to 2.7 MB. HTTP Archive also reports the median mobile home page reached 2.6 MB, up 8.4% from 2024. (HTTP Archive)

For business owners, this is where the web team has to be blunt. A tiny-looking widget can add another network request, another script, another consent issue, and another thing to debug when the form stops working.

Third-party script statistics: the hidden part of the stack

Third parties are where websites get messy. Analytics, ad pixels, heatmaps, chat tools, review widgets, consent platforms, social embeds, scheduling tools, and A/B testing tools all feel small when installed one at a time.

Together, they become the part of the website nobody owns.

18. HTTP Archive says pages with one or more third parties remain at 90% or higher. In plain English, nearly every modern website depends on code, content, or services served from outside the business’s own domain. (HTTP Archive)

19. HTTP Archive found low-ranked websites load more third-party requests overall. Across all sites, the median was 83 third-party requests on desktop and 79 on mobile. (HTTP Archive)

20. The top 1,000 sites had a median of 129 third-party requests on desktop and 106 on mobile. Bigger sites usually have more monetization, personalization, testing, analytics, and ad systems attached. (HTTP Archive)

21. Third-party requests increased year over year across all ranks. HTTP Archive reports the top 1,000 sites added 15 requests on desktop and 15 on mobile compared with 2024, while the broader dataset added five requests on each device type. (HTTP Archive)

22. The top third-party request categories in 2025 were ads, analytics, and CDNs. That is the commercial web in one sentence: sell, measure, and deliver. (HTTP Archive)

23. Third-party request types are dominated by scripts, images, and “other.” HTTP Archive says those three categories account for more than half of all third-party request content types. (HTTP Archive)

24. Google-owned services dominate the top 10 third-party domains by page count. HTTP Archive lists domains such as fonts.googleapis.com, googletagmanager.com, google-analytics.com, accounts.google.com, and adservice.google.com among the top entries, with facebook.com as the only non-Google domain in that top 10. (HTTP Archive)

If you run a business site, audit third parties like you audit expenses. Ask who owns each script, what it does, what happens if it fails, whether it fires on every page, and whether it still supports a current business goal.

Analytics and tag manager statistics

Measurement matters, but measurement bloat is real. A small business can go from “we need analytics” to “we have three analytics tools, two pixels, a heatmap, a chat widget, and a cookie banner blocking the hero” surprisingly fast.

25. W3Techs reports that 44% of websites use none of the traffic analysis tools it monitors. Some of those sites may use server-side analytics or tools outside the monitored list, but the number still shows how many websites have limited detectable measurement. (W3Techs)

26. Google Analytics is used by 44.9% of all websites. Among monitored traffic analysis tools, W3Techs gives Google Analytics an 80.2% market share. (W3Techs)

27. Meta Pixel is used by 8.9% of all websites. That makes it the second-largest monitored traffic analysis tool in W3Techs’ data. (W3Techs)

28. Microsoft Clarity is used by 3.8% of all websites. Clarity has become a normal small-business audit tool because it gives session recordings and heatmaps without the price tag of older enterprise UX tools. (W3Techs)

29. Hotjar is used by 1.9% of all websites. For conversion audits, that still makes it one of the more common behavior analytics tools in the monitored dataset. (W3Techs)

30. W3Techs reports that 53.8% of websites use none of the tag managers it monitors. That leaves a large share of websites where tags may be hard-coded, missing, or handled through platform integrations instead of a central container. (W3Techs)

31. Google Tag Manager is used by 46.0% of all websites. W3Techs gives it 99.7% market share among monitored tag managers, which is about as close to a category monopoly as web tech gets. (W3Techs)

The practical move is simple: document your measurement stack. A spreadsheet with script name, owner, purpose, pages loaded, consent category, and last-reviewed date can save hours when something breaks.

CDN and delivery statistics

Content delivery often gets treated as hosting plumbing, but it affects uptime, speed, security headers, caching, script dependencies, and how much control your team actually has.

32. W3Techs reports that 75.9% of websites use none of the JavaScript content delivery networks it monitors. Many sites still serve JavaScript from their own CMS, theme, plugin folder, app platform, or bundled build output. (W3Techs)

33. CDNJS is used by 11.7% of all websites. W3Techs gives CDNJS 48.5% market share among monitored JavaScript CDN usage. (W3Techs)

34. jsDelivr is used by 9.0% of all websites. It has 37.3% market share in W3Techs’ monitored JavaScript CDN category. (W3Techs)

35. Google Hosted Libraries is used by 7.5% of all websites. That puts it at 31.1% market share among monitored JavaScript content delivery networks. (W3Techs)

Here is the business question: if a CDN, tag manager, or external script has a bad day, does your website degrade gracefully, or does your quote form, navigation, checkout, or calendar stop working?

Performance statistics: the stack shows up in Core Web Vitals

Performance is where all the invisible decisions become visible. The CMS, theme, scripts, media, fonts, and third parties show up in how the page feels.

36. Google recommends an LCP of 2.5 seconds or less. Largest Contentful Paint measures loading performance, usually when the main content becomes visible. (Google Search Central)

37. Google recommends an INP under 200 milliseconds. Interaction to Next Paint measures responsiveness, which is where heavy JavaScript often hurts users. (Google Search Central)

38. Google recommends a CLS score below 0.1. Cumulative Layout Shift measures visual stability, which is why late-loading ads, fonts, banners, and embeds can hurt the experience. (Google Search Central)

39. HTTP Archive reports mobile Core Web Vitals improved from 36% in 2023 to 44% in 2024 and 48% in 2025. The web is improving, but the majority of mobile origins still do not pass all three metrics. (HTTP Archive)

40. Desktop Core Web Vitals improved from 48% in 2023 to 55% in 2024 and 56% in 2025. Desktop is better than mobile, but it also barely moved in the latest year of HTTP Archive’s data. (HTTP Archive)

41. Among the 1,000 most popular mobile sites, 51% had good Core Web Vitals. HTTP Archive found performance then dropped for the next popularity tiers before improving again among less popular sites. (HTTP Archive)

42. Mobile Core Web Vitals dropped to 42% for the next 10,000 sites and 37% for the next 100,000. HTTP Archive suggests mid-popularity sites may combine higher complexity with less sustained performance investment. (HTTP Archive)

43. Secondary pages beat home pages on Core Web Vitals by 14 percentage points on desktop and 11 points on mobile. HTTP Archive notes that secondary pages often benefit from cached resources and more consistent templates. (HTTP Archive)

That last stat is useful in client conversations. Home pages are often the slowest because everyone wants their section, widget, proof point, video, announcement, and animation above the fold.

AI crawler and bot statistics: a new pressure on the stack

AI crawlers changed the stack conversation. They don’t just raise content strategy questions. They raise technical questions about server load, robots.txt, rendering, raw HTML, structured content, and whether important information is visible without client-side JavaScript.

44. Cloudflare says global internet traffic grew 19% in 2025. More traffic doesn’t automatically mean more customers, especially when a rising share comes from bots, crawlers, agents, and automated clients. (Cloudflare)

45. Cloudflare handles more than 81 million HTTP requests per second on average and more than 129 million at peak. Its 2025 Radar report uses that network view to analyze traffic, bot, security, connectivity, and DNS patterns. (Cloudflare)

46. Googlebot was responsible for the highest volume of request traffic to Cloudflare in 2025. Cloudflare says Googlebot crawled customer sites for search indexing and AI training. (Cloudflare)

47. Googlebot accounted for more than a quarter of Verified Bot traffic. That is a reminder not to treat all bot traffic as junk. Some of it is tied to search visibility. (Cloudflare)

48. Cloudflare reported AI “user action” crawling increased by over 15x in 2025. This is the type of traffic site owners will increasingly need to distinguish from normal search crawls and human visits. (Cloudflare)

49. Non-Google AI bots accounted for 4.2% of HTML request traffic, while Googlebot alone accounted for 4.5%. This puts AI and search crawler traffic into a measurable operational category, not a theoretical future issue. (Cloudflare)

50. HTTP Archive notes that all key information should be present in initial raw HTML because AI crawlers do not render JavaScript. That means content hidden behind client-side rendering, tabs, delayed scripts, or broken hydration may be less visible to crawlers that matter. (HTTP Archive)

For small businesses, don’t overcomplicate this. Make your important service information, pricing cues, location details, FAQs, proof, and contact options visible in the HTML. If a customer, search crawler, or AI crawler needs a JavaScript miracle to understand the page, you’ve created unnecessary risk.

How to use these stats in a real website audit

Don’t turn this into a 90-page report nobody reads. Use the data to ask better questions.

  1. Inventory the stack. List the CMS, hosting, theme, plugins, forms, analytics, pixels, tag manager, cookie tool, CDN, major embeds, and critical third-party services.
  2. Cut what has no owner. If nobody can explain what a script does, why it loads, who uses the data, and what breaks if it is removed, it should be questioned.
  3. Measure the pages that make money. Test the home page, top service pages, landing pages, quote forms, checkout, booking page, and contact page with Core Web Vitals, not just a desktop Lighthouse score.
  4. Protect the basics. Keep key content in HTML, compress media, limit scripts on lead pages, document tracking, review consent behavior, and give every critical integration a fallback.

FAQ

What is a website tech stack?

A website tech stack is the set of systems, tools, and services that make a site work. It usually includes the CMS, hosting, frontend code, theme, plugins, analytics, pixels, forms, CDN, consent platform, security tools, and third-party integrations.

Why does a website tech stack matter for business owners?

The stack affects speed, security, maintenance cost, lead tracking, SEO, privacy compliance, and how fast your team can make changes. A messy stack can make a good-looking website expensive to run and hard to fix.

What should I audit first?

Start with the money pages: home page, top service pages, landing pages, booking pages, quote forms, checkout, and contact page. Then review every third-party script that loads on those pages.

Is WordPress still the most common CMS?

Yes. W3Techs reports WordPress is used by 41.5% of all websites and has 59.3% CMS market share, while HTTP Archive reports WordPress powers roughly 64% of CMS-driven sites. (W3Techs, HTTP Archive)

How often should a business review its website stack?

Review critical scripts and plugins quarterly, and run a deeper audit before a redesign, migration, ownership change, or major ad spend increase.

If your website stack has become hard to explain, hard to update, or hard to trust, we can help you clean it up. Start here and we’ll look at what’s actually running under the hood.