The median home page of 50 US-listed machinery and electrical equipment manufacturers we tested on 2026-10-21 transferred 5,933 KB on a mobile profile, 2.4 times the 2,504 KB median for home pages across the web in the HTTP Archive's September 2026 crawl. Most of the gap is autoplaying video, oversized photography and JavaScript, and the enterprise platforms these companies favour pass Core Web Vitals less often than the web as a whole.
No public dataset labels a website as a manufacturer's, so this page combines Chrome's real-user data for the platforms and tools industrial sites run, Google's 2017 industry benchmark and our own test of 50 home pages.
The thresholds, and why many manufacturers have no score
A page passes Core Web Vitals when 75% of real Chrome visits reach a Largest Contentful Paint (LCP) of 2.5 seconds or less, an Interaction to Next Paint (INP) of 200 milliseconds or less and a Cumulative Layout Shift (CLS) of 0.1 or less, judged separately for mobile and desktop (web.dev). Search Console reads the last 28 days and gives each group of URLs the status of its worst metric.
Traffic is the catch. The Chrome UX Report only includes pages and origins with an unpublished minimum of visitors, and it leaves out Chrome on iOS and Microsoft Edge. When a URL group lacks data, Search Console falls back to the origin; when the origin lacks it, nothing is shown. A product family page with a few hundred visits a month usually has no score of its own. Our B2B SaaS Website Speed and Core Web Vitals Statistics 2026 post covers the mechanics.
Google's industry benchmark dates from 2017, and industrial servers were the slowest
In January 2017 Google tested more than 900,000 mobile ad landing pages in 126 countries on an emulated phone over 3G and broke the results out by sector. It named "business and industrial markets", the closest category to manufacturing, among the three sectors with the most room to improve.
| Measure | Business and industrial markets | Range across 8 sectors | Google's best practice |
|---|---|---|---|
| Time to first byte | 2.7 s | 1.8 s to 2.7 s | Under 1.3 s |
| Speed Index | 8.7 s | 7.9 s to 11.3 s | Under 3 s |
| Page weight | 1.6 MB | 1.3 MB to 2.3 MB | Under 500 KB |
| Requests per page | 83 | 78 to 140 | Under 50 |
Source: Google, Find Out How You Stack Up to New Industry Benchmarks for Mobile Page Speed, February 2017
Industrial pages were lighter than most and still the slowest to answer in the United States. They were also slowest in Japan, at 1.9 seconds, though mid-table in the United Kingdom and Germany. The lab setup is long obsolete, so read the ranking, not the seconds.
Pass rates for the platforms manufacturers run on
The HTTP Archive detects the technology on each origin and joins it to Chrome's field data. The table covers the platforms we found on the manufacturers we tested, plus the open-source systems a mid-sized manufacturer is likely to be quoted. September 2026 is the latest month published.
| Platform | Mobile origins | All three, mobile | All three, desktop | Good TTFB, mobile | All three, mobile, top 1M | Median mobile JS (KB) |
|---|---|---|---|---|---|---|
| All origins | 8,569,446 | 52.8% | 60.8% | 45.9% | 46.6% | 714 |
| Adobe Experience Manager | 19,174 | 33.9% | 44.3% | 44.1% | 33.2% | 1,338 |
| Sitecore | 5,890 | 39.3% | 48.7% | 27.2% | 34.5% | 1,349 |
| Microsoft SharePoint | 2,067 | 26.2% | 39.8% | 32.8% | 25.8% | 1,244 |
| Progress Sitefinity | 1,925 | 55.7% | 58.1% | 38.9% | 50.1% | 1,159 |
| Optimizely CMS | 1,820 | 71.1% | 64.1% | 63.1% | 62.6% | 1,034 |
| Drupal | 93,193 | 64.7% | 66.1% | 45.6% | 58.6% | 650 |
| WordPress | 2,658,175 | 48.3% | 53.7% | 22.9% | 50.1% | 661 |
| TYPO3 | 40,300 | 80.3% | 72.3% | 59.6% | 74.4% | 307 |
| Pimcore | 2,331 | 79.5% | 72.3% | 57.5% | 71.1% | 659 |
| Umbraco | 4,332 | 76.4% | 73.0% | 74.1% | 70.9% | 927 |
| HubSpot CMS | 13,904 | 60.0% | 75.1% | 67.3% | 45.2% | 1,233 |
| Webflow | 59,483 | 71.3% | 82.6% | 73.1% | 49.8% | 904 |
Source: HTTP Archive Core Web Vitals and page-weight technology reports, all geographies, read through the public report API on 2026-10-21. KB is 1,024 bytes.
The enterprise suites trail the web. Adobe Experience Manager, Sitecore and SharePoint origins pass on mobile 26% to 39% of the time, 13 to 27 points below average, with close to twice the median page's JavaScript. They are the platforms large manufacturers buy.
Lighter stacks lead. TYPO3 origins pass at 80.3% with a median 307 KB of JavaScript, less than half the web's median. Pimcore and Umbraco sit just behind.
WordPress fails on the server. Only 22.9% of WordPress origins have a good time to first byte on mobile, the lowest in the table, which echoes Google's 2017 finding.
We build marketing sites in Webflow, so weigh that row with our bias in mind; its lead shrinks to 3.2 points among the top million origins. Adobe Experience Manager has barely moved since January 2024, from 32.7% to 33.9%, while the web rose from 46.8% to 52.8% (HTTP Archive).
What industrial sites add after launch
Industrial home pages also carry consent managers, video players, site search and image servers. The second column counts how many of our 50 home pages loaded each tool before cookies were accepted, so it is a floor.
| Tool | On our 50 home pages | Mobile origins | All three, mobile | All three, mobile, top 1M | Median mobile JS (KB) |
|---|---|---|---|---|---|
| All origins | Baseline | 8,569,446 | 52.8% | 46.6% | 714 |
| Google Tag Manager | 43 | 2,547,920 | 46.5% | 44.2% | 1,174 |
| OneTrust | 25 | 107,248 | 42.5% | 38.0% | 1,560 |
| YouTube embed | 4 | 720,396 | 49.3% | 48.6% | 1,564 |
| lite-youtube-embed facade | Not detectable | 7,333 | 50.5% | 53.0% | 834 |
| Vimeo | 4 | 167,785 | 48.9% | 43.8% | 1,179 |
| Google Maps | 4 | 823,804 | 50.6% | 46.6% | 988 |
| Coveo site search | 5 | 846 | 28.5% | 24.4% | 1,822 |
| Adobe Dynamic Media Classic (Scene7) | 4 | 4,419 | 26.0% | 20.7% | 875 |
Source: HTTP Archive Core Web Vitals and page-weight technology reports, read 2026-10-21; first column from Octo Agency's Lighthouse test of 2026-10-21
These are correlations, not tests: a site with OneTrust and Coveo is usually a large company with many other scripts.
Consent managers are the most common heavy script. OneTrust origins carry a median 1,560 KB of JavaScript and pass 10 points below average. On four of our 50 home pages the cookie banner's text was the LCP element for a first-time visitor, so the banner, not the product, is what Google times.
Video is cheaper behind a click. Origins using the lite-youtube-embed facade carry about half the JavaScript of those with the standard player. Google's Lighthouse guidance shows a 3 KB facade standing in for a 540 KB player until someone presses play.
Search and image servers cost responsiveness. Coveo and Adobe Dynamic Media Classic origins pass at 28.5% and 26.0%, and only 55.1% of Dynamic Media origins have a good INP on mobile, against 79.4% for all origins.
What 50 manufacturers' home pages load on a phone
We took every S&P 500 and S&P MidCap 400 company classified under industrial machinery, construction machinery and heavy trucks, agricultural machinery, or electrical components, dropped a distributor and a leasing company, and ran Lighthouse 12.8.2 once on each of the 52 home pages with its default mobile profile. Two could not be measured, one behind a bot challenge and one that refused the request, leaving 50.
| Measure | Median of 50 | Middle half of the 50 | Web median home page, September 2026 |
|---|---|---|---|
| Total transfer | 5,933 KB | 3,470 to 11,759 KB | 2,504 KB |
| JavaScript | 1,296 KB | 705 to 2,326 KB | 714 KB |
| Images | 1,541 KB | 614 to 2,991 KB | 753 KB |
| Requests | 81.5 | 60 to 152 | Not reported |
| Third-party services loaded before consent | 8.5 | 6 to 13 | Not reported |
Source: Octo Agency Lighthouse test, 2026-10-21, raw results in our research file; web medians from the HTTP Archive page-weight report
Video is the biggest single cost. 16 pages downloaded a video file while loading, a median 4.8 MB each, and on nine a video or its poster frame was the LCP element. Of the five heaviest pages, 20 MB to 48 MB each, four were heavy because of video or oversized images and one because of an 18 MB JavaScript file.
One image can outweigh a page. Nine pages served an image over 1 MB; the three largest weighed 31.6 MB, 11.5 MB and 9.0 MB.
Script is above the web median almost everywhere. 37 pages carried more JavaScript than the web's 714 KB median. Before any cookies were accepted, a tag manager loaded on 44, ad pixels on 27, an accessibility overlay on nine and session recording on six.
Platforms. Asset paths identified the platform on 41 of the 50: Adobe Experience Manager and Drupal on eight each, Sitecore on seven, WordPress on six and Next.js on four.
Requests have barely moved since 2017, a median of 81.5 against Google's average of 83, while weight has grown about 3.6 times. The methods differ, so read that as a direction.
Limits. These are large listed companies, not mid-sized firms, with one lab run per page. We tested from Serbia, so seven sites served a European or regional version and two a country selector. Lab timings depend on the test machine, so we report bytes and requests only.
What speed does to industrial leads
No published speed study covers manufacturing sites. The closest evidence is Google's, and it is old.
Deloitte and 55 for Google. Across 37 brands and more than 30 million sessions at the end of 2019, a 0.1 second improvement on four speed measures came with a 21.6% rise in lead-generation visitors progressing from the first form step to submission, and 5.5% more reaching the first step from the home page. It is correlational, but a multi-step RFQ form is the journey it measured.
Google and SOASTA, 2017. A model trained on bounce data found that as load time rose from one second to seven, the probability of a mobile bounce rose 113%.
Rankings. Google's documentation, updated September 2026, says Core Web Vitals "are used by our ranking systems" but that Search shows the most relevant page even when page experience is poor. On part-number searches, relevance wins first.
How to use this: a five-page speed test order for a manufacturing site
Field data shows which pages fail, not why. We would test in this order, which follows where buyers land and where a slow page costs a quote. The home page sits third, not first, although the Web Almanac's 2025 performance chapter found only 45% of mobile home pages passed, against 56% of secondary pages. The order complements the conversion test order in Manufacturing Website Conversion Rate Statistics 2026.
| Order | Page | Why it goes here | What to check first |
|---|---|---|---|
| 1 | Request for quote or contact page | Every lead passes through it, often with a drawing upload, a configurator or a sales router | INP on mobile with the form and its scripts loaded |
| 2 | Top product or product family page | Engineers land here from part-number and specification searches | LCP of the product image, spec table layout shift, CAD or 3D viewer loaded on click |
| 3 | Home page | The page most likely to fail, with the hero video, consent banner and most tags | Hero video weight, LCP element on first visit, tag count |
| 4 | Distributor or dealer locator | Map scripts and location data are heavy, and buyers use it on the road | Map loaded on interaction, INP while filtering |
| 5 | Top organic resource or datasheet page | Often the largest share of origin traffic, so it sets the origin score | Embedded video players, PDF previews, consent banner shift |
Source: Octo Agency recommendation, based on the HTTP Archive, Web Almanac and Lighthouse findings above
For each page, read the 28-day field data in Search Console or the Chrome UX Report first, then run a lab test on a throttled mobile profile to find the cause. Run it twice: once as a new visitor who sees the consent banner, and once with every script accepted, because many tags load only after consent.
All three datasets point the same way: industrial sites get slow one decision at a time, a hero video here, a 30 MB product shot and another pre-consent tag there. On a rebuild we would set a budget before the content arrives: an LCP of 2.5 seconds and an INP of 200 milliseconds on mobile for the five pages above, video loaded on click or as a compressed loop under 1 MB, no image over 300 KB, and a written list of every third-party script with an owner and a reason. Our marketing websites service page explains how we scope a manufacturer's site.
