M MurmTools

Meta on 276 sites, and what else was on the page

Meta is the company behind Facebook, Instagram and WhatsApp, and almost none of the places your browser meets it are those sites. It arrives on other people's pages as advertising and measurement code: the pixel a marketing team pastes into a checkout, an embedded post, a share or login button. This index attributes at least one contacted domain to Meta on 276 of the 5,294 sites measured — 5.2% of the corpus, and the eleventh-widest reach of the 88 companies broken out here. The more interesting number is who else was there.

It arrives with Google 92% of the time

Google was present on 254 of Meta's 276 sites. After that the list falls away quickly: LinkedIn on 71 sites (25.7%), Cloudflare on 40 (14.5%), Amazon CloudFront on 37 (13.4%), jsDelivr on 35 (12.7%), Microsoft Clarity on 28 (10.1%), Microsoft Ads on 27 (9.8%), OneTrust on 26 (9.4%), TikTok Pixel on 24 (8.7%) and Microsoft on 23 (8.3%).

Read as a group, that is a marketing department's shopping list rather than a technical stack: two or three ad platforms so campaigns can be attributed, a professional network for B2B retargeting, a content delivery network, a session analytics product, a consent tool. Meta is rarely the reason a page is instrumented. It is one line item on a page that was already instrumented.

87.3% early contact — low, for this index

On 241 of the 276 sites, at least one company was contacted before a consent banner could have been answered. That share ranks 79th of the 88 companies here, where rank 1 is the highest share, so only nine of the 88 sit lower.

The measurement is per site and it counts, not names: it records how many companies arrived inside the window, never which ones. So the supportable statement is that Meta was present on 241 sites where early contact happened, alongside a median of 6 companies that arrived in that window. On a page reaching 9 companies in total, that is a wide field, and picking a name out of it would be a guess.

The median page, and the one company that comes after

The median site carrying Meta contacted 9 companies over 49 requests, 3 of them classified as tracking against the vendor catalogue. Median time to first third-party contact was 793 ms, which ranks 22nd of 88 — the slower end. The fastest first contact recorded in the group was 55 ms.

The median number of companies that appeared after the consent window closed is 1. So the typical shape here is a page that reaches five companies before anyone could have answered anything, and adds one more afterwards.

77 of 276 had a banner at all

Consent software was detected on 77 sites, a little over a quarter. OneTrust was seen alongside on 37 of them, the IAB TCF framework on 31, Cookiebot on 4, and Osano, CookieYes and Klaro on 1 each. These are sightings, not slices: a site can carry more than one banner.

The other 199 sites presented nothing to answer. That is the ordinary condition of a Meta-carrying page in this corpus, and it is why the early-contact figure needs reading carefully rather than loudly — on most of these sites there was no consent moment to be early to.

Twelve sites, named

harvard.edu contacted 5 companies, all 5 inside the window, first contact at 272 ms, no banner detected. scribd.com: 12 and 12, behind an Osano banner. hotjar.com: 18 and 18, behind OneTrust. media.net: 8 and 8, behind an IAB TCF banner. zillow.com: 10 with 9 inside the window.

Then the ones that waited. dell.com contacted 15 companies with 4 inside the window, first contact at 1,352 ms behind OneTrust. t-mobile.com: 26 with 6, at 951 ms. kaspersky.com: 22 with 3, at 1,460 ms. oxylabs.io: 12 with 2, at 857 ms behind a Cookiebot banner. shopee.co.id: 4 with 3. vungle.com reached 14 companies and none of them inside the window, first contact at 2,116 ms.

Meta's own site is in the sample. meta.com contacted 2 companies, both inside the window, first contact at 1,473 ms, with no consent banner detected.

What this page can and cannot tell you

It can tell you how often Meta code turns up on other people's sites in this sample, how heavy those sites are, and which other companies keep it company. It cannot tell you what was sent, because request payloads were never inspected, and it cannot tell you whether any particular visit was linked to an account.

If you want the same reading for a page you actually care about — your employer's site, your own checkout — MurmStack runs the equivalent check on whatever is open in the browser and lists the companies contacted, in order, with the pre-consent window marked.

Highest-ranked sites where this was seen
SiteCompaniesBefore consentFirst contact
kaspersky.com2231460 ms
vungle.com1402116 ms
oxylabs.io122857 ms
harvard.edu55272 ms
media.net88745 ms
dell.com1541352 ms
t-mobile.com266951 ms
scribd.com1212255 ms
hotjar.com1818757 ms
meta.com221473 ms
zillow.com109679 ms
shopee.co.id431292 ms
Consent tools seen on these sites
ToolSites
OneTrust37
IAB TCF consent tool31
Cookiebot4
Osano1
CookieYes1
Klaro1

The six most common are listed. The remaining 2 of 77 sites with a banner carry something outside that six. One tool is recorded per site, so these counts never overlap.

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

What counts as Meta being present on a site?

The browser opened a connection to a domain this index attributes to Meta during a single cold page load. That covers advertising and measurement code as well as embeds; it does not distinguish between them, and it says nothing about what the request carried.

Did the Meta pixel fire before consent on these sites?

This corpus cannot answer that. Early contact is recorded per site as a count of companies arriving in the window, not as a list of names, so the honest statement is that Meta was present on 241 sites where such contact happened.

Why is Google on 92% of Meta's sites?

Because a page that installs one advertising platform generally installs several, and Google is the company most often encountered across these 276 pages. The figure describes co-presence on the same page and nothing about a relationship between the two companies.

Is 276 sites a lot?

It is 5.2% of the 5,294 sites measured, and the eleventh-widest reach of the 88 companies this index breaks out. Ten companies here appear on more sites than Meta does.

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