M MurmTools

WordPress.com: light pages that reach out early

WordPress.com is the hosted version of WordPress, run by Automattic — the version where the platform manages the servers rather than the site owner. The entry here covers the hosts a page contacts when it is served by that platform or uses its shared services, and the index classifies it as a CMS rather than a tracking service. It was seen on 59 of the 5,294 sites measured. Its profile is an unusual combination: the lightest pages in this index also make the earliest first contact in it.

381 ms, and nothing in the index lower

The median site here makes its first third-party contact 381 ms into the page load. Of the 88 companies in this index, none records a lower median — this entry sits at rank 88 of 88 when they are ordered from latest to earliest. The fastest single site in the sample was 80 ms. Two of the named examples, akismet.com at 142 ms and automattic.com at 144 ms, land in the same region.

And the pages are nearly the lightest too

The median site makes 21 requests, which is 86th of 88 — only two entries here are lower. It contacts 6 companies (78th of 88) and carries 1 company classified as tracking (66th of 88). That combination is worth sitting with, because the intuitive story runs the other way: heavy pages are usually assumed to be the ones that reach out fastest. Here the lightest pages do. A plausible mechanical reading is that when a page has few requests to make, the third-party one is not queued behind anything — but this index measures order and timing, not the reason for them, so treat that as a hypothesis rather than a finding.

Early contact on 56 of 59

On 56 sites — 94.9 percent — at least one company was contacted before a consent banner could have been answered, ranking 35th of 88 on share. The median count of companies inside that window is 5 (57th of 88, level with 22 others), and the median count contacted after it closes is 0. That last figure is consistent with the timing: on pages this small, the sequence is finished early.

Banners on 22 of 59

A consent banner was detected on 22 sites. Seen alongside: OneTrust on 11, an IAB TCF consent tool on 6, Usercentrics on 2, and Osano, CookieYes and Cookiebot on 1 each. Sites can carry more than one, so these are separate counts. Elsewhere on the same pages, Google is present on 62.7 percent, Parse.ly on 25.4 — a publisher-analytics tool, which fits the sample — OneTrust on 18.6, Cloudflare on 15.3, HubSpot and Adobe Fonts on 11.9 each.

Blogs, publishers and institutions

The sample mixes platform properties with the sites built on them: wordpress.com, gravatar.com, tumblr.com, wp.com, akismet.com and automattic.com, alongside harvard.edu, pewresearch.org, ssrn.com, cnet.com, nypost.com and stackadapt.com. pewresearch.org contacted 9 companies with all 9 inside the window at 173 ms, and harvard.edu 5 with all 5 inside it at 272 ms — both light pages that start immediately. cnet.com contacted 13 with 12 inside the window. nypost.com is the counter-example: 8 companies, none inside the window, first contact at 2,882 ms behind a OneTrust banner.

Highest-ranked sites where this was seen
SiteCompaniesBefore consentFirst contact
wordpress.com33303 ms
gravatar.com22711 ms
tumblr.com22770 ms
harvard.edu55272 ms
wp.com33865 ms
cnet.com1312459 ms
stackadapt.com951351 ms
nypost.com802882 ms
akismet.com22142 ms
pewresearch.org99173 ms
automattic.com33144 ms
ssrn.com42634 ms
Consent tools seen on these sites
ToolSites
OneTrust11
IAB TCF consent tool6
Usercentrics2
Osano1
CookieYes1
Cookiebot1

Seen on the same pages

Companies most often contacted by the same sites. Co-presence only — the crawl records what arrived, not who called whom.

Questions

Is WordPress.com counted as a tracker in this index?

No. It is catalogued as a CMS and not classified as tracking, so its presence does not add to the tracker counts on the sites where it appears.

Why do these pages make so few requests?

The median of 21 requests describes the sites in the sample, which include a lot of text-first blogs, institutional pages and platform properties. Nothing in the measurement attributes the low count to the platform itself.

Does an early first contact matter if the page is light?

It matters for what happens before a banner can be answered: on 56 of the 59 sites, at least one company was contacted inside that window. A small page still makes that contact, it just makes fewer afterwards.

Is this the same as the WordPress entry in the technology index?

No. This entry counts sites that contacted WordPress.com-operated hosts during page load. The technology index records the software a page was built with, which is a different measurement on a different set of sites.

Check any page yourself MurmStack is the tool that produced these figures. Free, two permissions, no host access.
Add to Chrome — free