The median home page of the 26 healthcare software and data vendors we tested on 21 October 2026 transferred 3,839 KB on a simulated mid-range phone and carried 1,347 KB of JavaScript. The median page among the top million origins in the HTTP Archive's September 2026 data weighs 2,670 KB with 904 KB of JavaScript, so the vendor pages were about 1.5 times heavier on both counts. Six of the 26 passed 10 MB, and in five of those six the weight was video or oversized images, not code.
No public dataset labels a website as healthcare B2B, so this page combines our own lab test of vendor home pages with Chrome's real-user data for the platforms and tools those pages run on, read through the HTTP Archive. The generic software numbers are in our B2B SaaS Website Speed and Core Web Vitals Statistics 2026 post.
Why there is no healthcare number in Google's data
Google judges a page on three field measures from real Chrome users. Largest Contentful Paint (LCP) should be 2.5 seconds or less, Interaction to Next Paint (INP) 200 milliseconds or less and Cumulative Layout Shift (CLS) 0.1 or less, each at the 75th percentile of visits over 28 days, separately on mobile and desktop (Google Search Console Help). A group of pages takes the status of its worst metric.
The Chrome UX Report behind those scores has no industry field. It also needs a minimum number of visitors per page or origin, a number Google does not disclose, and it leaves out Chrome on iOS, Android WebView and other Chromium browsers such as Microsoft Edge (CrUX methodology). A clinical director reading a vendor site on an iPhone, or a hospital IT reviewer on a managed Edge browser, is not counted. When a group of pages lacks data, Search Console falls back to the origin, and when the origin lacks it, the site does not appear at all.
So we built the healthcare view from two halves. The first is a lab test: Lighthouse 12.8.2 on its default mobile profile, two runs per home page, taking the median, on 26 established vendors that sell software or data to providers, payers and life sciences firms. We picked them by hand, so treat them as a snapshot of the category, not a random sample. R1 RCM blocked the test with a 403 on both runs and is excluded. The second is field data: for each platform and tool found on those pages, the HTTP Archive's technology report gives the share of origins passing Core Web Vitals.
The platforms healthcare vendors build on, and how those platforms score in the field
WordPress ran 13 of the 26 home pages and Webflow six. Next.js ran three, and Drupal, Sitecore and HubSpot CMS one each; athenahealth runs a custom build. We read September 2026 field data for each platform among the top million origins by Chrome traffic, the closest match to an established vendor's site.
| Platform | Vendor home pages | Top 1M origins | All three, mobile | All three, desktop | Good LCP, mobile |
|---|---|---|---|---|---|
| All top 1M origins | n/a | 747,521 | 46.6% | 61.9% | 69.7% |
| WordPress | 13 | 110,537 | 50.1% | 55.8% | 67.2% |
| Webflow | 6 | 2,635 | 49.8% | 72.3% | 64.9% |
| Next.js | 3 | 39,764 | 25.5% | 49.3% | 53.8% |
| Drupal | 1 | 11,658 | 58.6% | 62.5% | 75.9% |
| Sitecore | 1 | 1,732 | 34.5% | 51.3% | 53.8% |
| HubSpot CMS | 1 | 704 | 45.2% | 64.8% | 67.5% |
Source: Octo Agency platform detection, 21 October 2026; HTTP Archive Core Web Vitals technology report, rank filter Top 1M, September 2026, read 2026-10-21
The platform gap is small on busy sites. WordPress and Webflow sit within four points of the top-million average on mobile. Webflow leads on desktop. We build marketing sites in Webflow and applications in Next.js, so we are not neutral on either row. The Next.js figure covers web applications and shops as well as marketing sites, and a server-rendered Next.js site with little script can pass.
Drupal is the outlier worth knowing. Its top-million origins passed at 58.6% on mobile, 12 points above average. Only one vendor in our sample, Definitive Healthcare, uses it.
The tools on the page cost more than the platform
Every page in the sample carried something added after launch. We counted a tool when the home page HTML referenced it or the browser loaded it during the test. Tools that load only after consent are undercounted, because Lighthouse does not accept the banner.
| Tool | Vendor home pages | All three, mobile | Good INP, mobile | Median mobile JavaScript (KB) |
|---|---|---|---|---|
| All top 1M origins | n/a | 46.6% | 65.7% | 904 |
| Google Tag Manager | 20 | 44.2% | 66.5% | 1,418 |
| LinkedIn Insight Tag | 12 | 32.1% | 53.3% | 1,972 |
| OneTrust | 10 | 38.0% | 62.9% | 1,784 |
| HubSpot tracking and forms | 10 | 36.7% | 59.4% | 1,702 |
| Marketo | 8 | 33.2% | 56.6% | 1,948 |
| Wistia | 8 | 46.1% | 68.2% | 2,012 |
| Vimeo | 7 | 43.8% | 75.7% | 1,671 |
| Qualified | 7 | 12.6% | 33.0% | 2,087 |
| ZoomInfo | 5 | 33.8% | 57.8% | 1,912 |
| 6sense | 5 | 21.8% | 44.5% | 2,138 |
| Osano | 3 | 44.6% | 70.7% | 1,334 |
Source: Octo Agency tool detection, 21 October 2026; HTTP Archive Core Web Vitals and page-weight technology reports, rank filter Top 1M, September 2026. KB is 1,024 bytes.
These are correlations. A company that buys 6sense and Qualified usually runs a long list of other scripts too, and the table cannot isolate one tool. Still, every sales and marketing tool on the list sits below the average, and the two account-based sales tools, Qualified and 6sense, sit lowest. INP falls furthest, because script that runs at load competes with the visitor's first tap.
Our B2B SaaS speed post has the wider tool list. The healthcare difference is the consent layer: 15 of 26 vendor pages ran OneTrust, Osano, TrustArc or Cookiebot, a sign that privacy review shapes these sites. Our Healthcare B2B Landing Page Statistics 2026 post covers what HHS tracking guidance still requires after the 2024 court ruling.
What our lab test found on each vendor home page
The table sorts the 26 pages by weight. Requests and timings are the median of two runs.
| Vendor | What it sells | Platform | Transfer (KB) | JavaScript (KB) | Requests | Lab LCP (s) | Lab CLS |
|---|---|---|---|---|---|---|---|
| Artera | Patient communications | WordPress | 21,389 | 1,982 | 225 | 15.6 | 0.035 |
| Waystar | Revenue cycle software | WordPress | 18,299 | 2,781 | 192 | 25.8 | 0.010 |
| Innovaccer | Healthcare data platform | Webflow | 18,291 | 1,197 | 146 | 86.8 | 0.032 |
| Phreesia | Patient intake software | WordPress | 17,510 | 2,417 | 197 | 16.5 | 0.166 |
| Flatiron Health | Oncology data and software | HubSpot CMS | 13,649 | 312 | 57 | 14.3 | 0.000 |
| Tebra | Practice software | Next.js | 10,091 | 3,396 | 225 | 40.6 | 0.000 |
| WellSky | Post-acute care software | WordPress | 7,642 | 2,466 | 239 | 12.1 | 0.009 |
| Cedar | Patient billing | Webflow | 6,102 | 1,582 | 139 | 11.0 | 0.003 |
| athenahealth | EHR and practice software | Custom | 5,728 | 3,529 | 360 | 26.2 | 0.116 |
| Tempus | Precision medicine | Next.js | 4,134 | 2,421 | 142 | 20.8 | 0.908 |
| Medidata | Clinical trial software | WordPress | 4,095 | 1,227 | 206 | 12.0 | 0.000 |
| Commure | Health system software | Webflow | 4,070 | 428 | 100 | 8.8 | 0.014 |
| Health Catalyst | Healthcare analytics | Webflow | 3,924 | 1,191 | 62 | 9.3 | 0.010 |
| Availity | Payer and provider network | WordPress | 3,754 | 1,585 | 105 | 10.0 | 0.095 |
| NextGen Healthcare | EHR | Sitecore | 3,516 | 1,276 | 161 | 8.5 | 0.000 |
| Qventus | Hospital operations software | WordPress | 3,403 | 778 | 179 | 22.2 | 0.995 |
| ModMed | Specialty EHR | WordPress | 3,365 | 3 | 28 | 10.9 | 0.000 |
| Definitive Healthcare | Healthcare market data | Drupal | 2,967 | 1,686 | 144 | 12.5 | 0.173 |
| eClinicalWorks | EHR | WordPress | 2,446 | 701 | 113 | 11.4 | 0.089 |
| PointClickCare | Post-acute care EHR | WordPress | 2,348 | 1,926 | 119 | 4.0 | 0.012 |
| LeanTaaS | Capacity management software | WordPress | 2,342 | 1,630 | 209 | 13.9 | 0.046 |
| Abridge | Clinical documentation AI | Webflow | 2,231 | 1,417 | 110 | 10.1 | 0.000 |
| Veeva | Life sciences software | WordPress | 1,790 | 674 | 86 | 9.6 | 0.014 |
| Notable | Healthcare automation | Webflow | 1,684 | 1,170 | 74 | 5.8 | 0.002 |
| Komodo Health | Healthcare data | WordPress | 1,334 | 941 | 64 | 2.8 | 0.000 |
| Epic | EHR | Next.js | 1,013 | 308 | 49 | 6.1 | 0.000 |
Source: Octo Agency Lighthouse 12.8.2 tests, default mobile profile, two runs per page, 21 October 2026
Media, not code, makes the heaviest pages. Artera's home page loaded four MP4 animations of 2 to 3 MB each as it opened. Phreesia and Waystar loaded Vimeo video files before any click. Innovaccer served PNG images of up to 3,973 KB, and Flatiron Health a 9,676 KB PNG and a 7,015 KB animated GIF. Converting those files to compressed WebP or video would remove most of the weight without a design change.
Light pages exist on every platform. Epic's Next.js page transferred 1,013 KB and contacted no third-party host at all. Komodo Health's WordPress page scored 92, the only one in the 90 to 100 band that Lighthouse calls good, and reached its largest paint in 2.8 seconds. ModMed's page holds its scripts back with a non-standard script type, so the test saw 3 KB of JavaScript.
Layout shift has two sources. Five of 26 pages crossed the 0.1 CLS line. On Tempus, the shift came almost entirely from an AI assistant panel loading over the page, and its largest element was the cookie notice. On Qventus, a statistics section moved after load.
Consent banners and chat widgets: the two healthcare habits to manage
Consent banners. Google's guidance says cookie notices are "a very common source of layout shifts", that the Accept button often causes poor INP because accepting loads many scripts at once, and that a large notice can become the LCP element on mobile. Its advice: load the notice script asynchronously and directly in the page rather than through a tag manager, preconnect to its host, and reserve space or use an overlay so nothing moves (web.dev, 2024).
Chat and meeting widgets. Qualified's messenger script alone was 840 KB on Phreesia, Tebra and WellSky. Lighthouse's facade guidance names chat widgets such as Intercom and Drift, and the YouTube and Vimeo players, as candidates for a static placeholder that loads the real script on click; its YouTube example swaps a 540 KB player for 3 KB. That audit was removed in Lighthouse 13, but the technique still applies.
Third parties in general. Our median vendor page made 140 requests and contacted 14 third-party companies, counting Webflow's and HubSpot's file hosting. The Web Almanac 2025 puts the median mobile page at 79 third-party requests, with about 90% of pages loading at least one third party.
What speed does to rankings and leads
Google says Core Web Vitals "are used by our ranking systems", that "there is no single signal" for page experience, and that it shows the most relevant content even when page experience is poor (Google Search Central, updated September 2026). For a vendor competing on a category term, speed breaks ties between equally useful pages.
The lead evidence is old. A Deloitte and 55 study for Google followed 37 brands and more than 30 million sessions at the end of 2019. On lead-generation sites, a 0.1 second gain across four speed measures came with a 21.6% rise in visitors moving from the first form step to submission. It is correlational, and it predates INP. Our Website Speed Statistics 2026 post covers the ecommerce tests.
What this means for a healthcare B2B site: a five-page speed test order
Field data tells you which pages fail, and the lab tells you why. We would test in this order, because it follows where hospital, payer and pharma buyers land and where a slow page costs a lead.
| Order | Page | Why it goes here | What to check first |
|---|---|---|---|
| 1 | Demo or contact page | Every lead passes through it, with form, scheduler and chat scripts | INP on mobile with consent accepted |
| 2 | Home page | 45% of mobile home pages pass, against 56% of inner pages | Hero video and image weight, LCP element |
| 3 | Main buyer solutions page, such as health systems or payers | Paid and organic buyers land here | Video embeds, carousels, layout shift |
| 4 | Security, compliance or integrations page | Hospital IT and procurement reviewers read it during evaluation | Embedded documents and widgets, load on Edge and Safari |
| 5 | Top organic resource or report page | Often the largest share of origin traffic, so it sets the origin score | Consent banner shift, form embeds |
Source: Octo Agency recommendation, based on the lab and HTTP Archive findings above and the Web Almanac 2025
Run each lab test twice, once as a new visitor with the consent banner showing and once with every script accepted. Lighthouse's own variability guide recommends the median of five runs on a quiet machine; our two-run medians on a shared machine are good enough for weight and request counts, less so for timings.
Then set a budget before the next tool arrives. On our four-week marketing-site process we would write the script list into the brief, with an owner and a reason for each tool, and agree an LCP of 2.5 seconds and an INP of 200 milliseconds on mobile for the five pages above. Our marketing websites service page explains how we scope a build.
