Opens in a new tab
Home/Blog/How much data do ad blockers cost a WordPress site?

How much data do ad blockers cost a WordPress site?

Privacy & CompliancePhillip StemannNovember 16, 2026

Last updated: November 18, 2026

Title card: How much data do ad blockers cost a WordPress site?
In this article
Key takeaways
  • There is no single “ad blocker loss” number. The best measurement I could verify is Plausible’s own 2021 study of one Hacker News and Reddit spike, where Google Analytics missed 58.67% of visitors. Plausible itself calls that an extreme case.
  • Blockers match requests, not intentions. Public filter lists such as EasyPrivacy (checked 2 October 2026) block googletagmanager.com outright and google-analytics.com script and XHR requests from third-party contexts, plus generic paths like /fbevents.js and /gtm.js.
  • Blockers cannot see what your server logs or what a vendor’s server sends to another vendor. They also cannot recreate a visit that your page-view tag never reported.
  • Server-side forwarding recovers conversion events (an order, a lead). It does not by itself recover the browsing session in a client-side tool, and it does not bypass consent.
  • Measure your own loss before spending anything: compare server logs or a second tool against GA4, check for blocked requests in your own browser, and read Search Console against organic sessions with the caveats below.

“Ad blockers kill 30% of your analytics” is the kind of sentence that gets repeated because it sounds plausible and is hard to check. It mixes two different things: how many people run an ad blocker, and how many of your page views never reach your analytics tool. They are related, but they are not the same number, and the second one depends heavily on who your visitors are. This post separates the two, shows what blockers actually match, and gives you a way to measure the gap on your own WordPress site. It belongs to our privacy and browsers series; the hub is Consent Mode v2 on WordPress.

What the usage numbers say, and what they do not

Ad-blocker usage figures are widely repeated, often from secondary sites that do not link the original survey. I only use a number that I could read at its source.

The figure I could read is from eyeo, the company behind Adblock Plus, in its 2026 ad-blocking report. It estimates about 1.1 billion global ad-blocking users in Q4 2025 and says they represent 30% of the total internet population. Two cautions. eyeo sells ad-filtering technology, so it is an interested party. And its user counts are internal estimates built from vendor and public data, which it describes in its methodology.

The same report has a 2025 country table (the chart caption cites the GWI Consumer Insights Platform, while the methodology section describes traffic data from Blockthrough’s bait-URL detection, so I treat the table as eyeo’s reported rates, not as a clean GWI quote). A few rows:

  • Denmark 23%, UK 27%, Germany 31%, USA 33%, Indonesia 41%, Japan 15%.

Even with the labelling caveat, the point holds: the share of people running some blocker is roughly a quarter to a third in many large markets, and it differs by country. What it does not tell you is how many of your page views vanish from GA4. A person with an ad blocker may still allow your analytics tag. A person with no ad blocker may use Firefox or Brave, which have their own tracker protection.

Two bar charts. Left: ad-blocking rate by country from eyeo's 2026 report, from Japan 15 percent to Indonesia 41 percent. Right: share of visitors Google Analytics missed by browser in Plausible's 2021 study, from Safari 41.4 percent to Firefox 88.3 percent.
Two different measurements, not to be added together or compared directly. Sources: eyeo Ad-blocking report 2026 (country table, 2025); Plausible, “58% of Hacker News, Reddit and tech-savvy audiences block Google Analytics” (August 2021 data, page updated May 2026).

The measurement that comes closest: analytics against analytics

The most useful public data compares two trackers on the same traffic. In August 2021 Plausible compared its own tool with Universal Analytics (the predecessor of GA4) on a site with a post trending on Hacker News and Reddit, over three days. Plausible was installed through a proxy so that it would not be blocked; Google Analytics was installed the default way. Per Plausible’s write-up (last updated 12 May 2026):

Metric Plausible Google Analytics Gap
Visitors 50,947 21,054 58.67%
Pageviews 60,731 25,275 58.38%
Laptop/desktop visitors 25,725 8,181 68.2%
Mobile/tablet visitors 27,331 12,848 49.9%
Firefox visitors 15,192 1,780 88.28%
Chrome visitors 16,405 8,133 50.42%
Safari visitors 16,136 9,454 41.41%

This is Plausible’s claim about its own test, and Plausible sells a competing product. It is also three days, one site, one extreme audience, and a measurement made before GA4 existed. Plausible says so itself: in its more recent post on whether ad blockers block Plausible, it calls the result “an extreme case”. In the first post it also reports an earlier study in which fewer than 10% of visitors blocked Google Analytics on foodie and lifestyle sites and more than 25% did on tech-focused sites. I did not re-fetch that older study, so treat those two figures as Plausible’s statement.

Plausible also flags that the gap is not all blockers: the two tools count unique visitors differently, and bot filtering differs. Only pageviews are comparable, and those showed almost the same gap as visitors.

What you can take from it: the difference between audiences is large, Firefox visitors vanished from GA far more than Safari visitors, and desktop is hit harder than mobile. What you cannot take from it: a number for your site.

What a blocker actually blocks

An ad blocker is a request filter. It compares each URL the page tries to load against rules and cancels the ones that match. The rules come from public lists. uBlock Origin ships with its own lists enabled by default plus EasyList and EasyPrivacy; I checked the uBlock Origin assets file and both are listed without an “off” flag. So the list contents are a fair proxy for what many desktop users run.

Here is what the lists contain as of 2 October 2026. These are rules I read in the files, not a test of each blocker’s behaviour.

# EasyPrivacy (easylist.to/easylist/easyprivacy.txt, version 202610021610)
||googletagmanager.com^
||google-analytics.com^$script,third-party,xmlhttprequest
||analytics.google.com^$third-party
||connect.facebook.net^*/fbds.js$third-party
/fbevents.js
/gtag/js?
/gtm.js
.facebook.com/tr?$image

# uBlock filters - Privacy (uBlockOrigin/uAssets filters/privacy.txt)
||google-analytics.com/analytics.js$script,xhr,redirect=google-analytics_analytics.js:5
||googletagmanager.com/gtag/js$script,xhr,redirect=googletagmanager_gtm.js:5
||googletagmanager.com/gtm.js$script,redirect=googletagmanager_gtm.js:5,domain=~nerc.com

Reading those rules:

  • Domain rules (||googletagmanager.com^) block everything from that host, whatever the path.
  • Path rules (/gtm.js, /fbevents.js) match the file name on any host. This matters: serving a tag from your own domain does not automatically hide a file called gtm.js.
  • Redirect rules in uBlock’s own list do not just fail the request. They serve a neutered stand-in script so that pages calling ga() do not throw errors. The page runs; nothing is sent.
  • Exceptions exist. EasyPrivacy has a handful of per-site exception lines, such as one for google-analytics.com/mp/collect? on a single named site. Do not read anything general into them.

The same list also contains ||plausible.io^$third-party and plausible script path rules, so Plausible’s statement that uBlock Origin’s own Privacy list has no Plausible entry is true of that one list but not of EasyPrivacy. No analytics vendor’s default script is immune from every list.

Browsers differ. Brave states on its Shields page that it blocks trackers and cross-site cookies by default. MDN’s Firefox tracking protection guide says Firefox ships a list of cross-site tracking sites and blocks content from them when tracking protection is on, naming analytics sites as common examples, and notes that your site then cannot rely on such services. Which Firefox mode applies by default is something I could not confirm from the pages I could read, so check Mozilla’s own current docs. Safari’s separate cookie rules are in our Safari ITP post.

Diagram of three request paths from a WordPress visitor: a browser tag to googletagmanager.com blocked by filter lists, a first-party server log that is not blocked, and a server-to-platform call that a browser blocker never sees.
What a browser-side filter can see, and what it cannot. Rule examples from EasyPrivacy and uBlock filters, checked 2 October 2026.

What blockers do not block

A blocker works inside the visitor’s browser (or DNS resolver, or network). Three things sit outside it.

  1. Your own server log. The visitor’s browser asked your WordPress site for a page. That request was to your domain and was necessary to show the page. No filter list blocks the HTML document of the site it is serving. Your web server or CDN logged it, bots included.
  2. Calls made server to server. If your server, or a vendor’s server, sends an event to Google, Meta or another platform, no browser is in the middle. A blocker in the visitor’s browser cannot cancel it.
  3. Requests that match no rule. Lists are community-maintained and can lag, so a request on a new endpoint or an unusual first-party path may go through. That is luck, not design. Plausible’s proxy advice is a version of this: it moves requests to a first-party URL, which in Plausible’s words “bypasses most blockers that target third-party tracking domains and known tracking URLs” (its claim, from its own post). Path-based rules such as /gtm.js can still match, and the same post says visitors who block JavaScript or your own domain remain uncounted.

Why the effect varies so much

The same site tag can lose 10% on one audience and more than half on another. The drivers are visible in Plausible’s data and in the lists:

  • Audience. Developers, security-minded readers and gamers install blockers and privacy browsers more often. Plausible’s numbers for a Hacker News spike are an upper edge, not a typical WordPress site.
  • Device. Desktop beat mobile in the same study (68.2% against 49.9% missing). Plausible attributes this partly to how hard it is to install a blocker in mobile Chrome.
  • Browser. Firefox lost 88% of its visitors in that test; Safari 41%. Your browser mix drives your loss.
  • Country. In eyeo’s table the ad-blocking rate runs from 15% in Japan to 41% in Indonesia. Denmark is listed at 23%.

For a typical WordPress business site with consumer visitors, expect a result between “barely noticeable” and “one in four”, and measure it rather than guess. For a developer-tools blog, plan for far more.

How to measure your own loss

Five methods, from most to least reliable. Each has a caveat, and the caveats matter more than the arithmetic.

1. Server logs against GA4

Ask your host or CDN for access logs and count requests to HTML pages with a 200 response, excluding obvious bots, your own IPs and admin paths. Compare that with GA4 pageviews for the same dates.

gap = 1 - (GA4 page_view events / human page requests in server logs)

Example (made up for illustration):
  server log, humans, HTML 200:  12,000
  GA4 page_view:                  9,000
  gap = 25%

That 25% is the combined effect of blockers, declined consent, cached pages that never fire the tag, visitors who leave before the tag loads, and bots that slipped through the log filter. It is an upper bound on blocker loss, not the blocker loss itself. If you use a full-page cache or a CDN, make sure your logs come from the layer that sees every request. Our post on why WordPress analytics data goes missing lists the other leaks.

2. A second analytics tool

Run a second tool, ideally one with a different script host or a proxy, for a few weeks and compare pageviews, not visitors, since tools count unique visitors differently (Plausible makes the same point). A gap is a hint, not a diagnosis: the second tool may also be blocked.

3. Search Console clicks against organic sessions

Search Console counts clicks from Google results to your pages. Compare those with organic search sessions in your analytics for the same URLs and dates. If clicks are consistently well above sessions on your top landing pages, some of the difference may be blocking. But clicks and sessions are not the same unit. Google’s Search Console performance documentation sets its own rules for what counts as a click (for example, a link that stays inside Google’s own platform is not counted), and sessions are defined differently again. Other differences, such as non-HTML files that carry no page tag, are my own reasoning rather than something that page states. Use this method only for direction and trend, and pick landing pages that are plain HTML.

4. Look at your own blocked requests

Install a blocker in a test browser profile, open your site with DevTools open and the Network tab filtered to gtag or collect. In Chromium, a request cancelled by an extension appears with net::ERR_BLOCKED_BY_CLIENT, which the Chromium network error list describes as “the client chose to block the request”. This does not give you a rate. It tells you whether the tags on your site are matched at all.

Checklist of five steps to measure ad blocker data loss on a WordPress site: pick a period, count human page requests in server logs, compare to GA4 page views, compare browser mix, and inspect blocked requests in DevTools.
A measure-your-own-loss checklist. Treat the result as an upper bound until you rule out consent and caching.

5. Compare by browser

Break your analytics down by browser and operating system and compare it with what you would expect. If Firefox is 3% of your sessions but you know that your audience skews technical, that is a sign. In Plausible’s test, Firefox was hit hardest. This is circumstantial evidence only.

What server-side forwarding does and does not recover

Server-side tracking moves one hop, the delivery to the platform, off the visitor’s browser. It helps with conversions. It does not fix everything.

  • Recovered: conversion events. If WordPress records an order or a form submission on the server, that event can be sent on to GA4, Meta and other platforms without relying on the visitor’s pixel. A blocked Meta pixel no longer means a lost Meta purchase. See WooCommerce server-side tracking.
  • Not recovered: the browsing session. If the page-view tag was blocked, GA4 has no session for that visitor. Google’s Measurement Protocol documentation says its intent is “to augment automatic collection through gtag, Tag Manager, and Google Analytics for Firebase, not to replace it”, and that events are joined to tagging data using client_id. Sent without a matching client-side session, a purchase can show up as an event but without the source, device and pages that tagging would have supplied. Exactly how it appears in your reports depends on what identifiers you send; test it.
  • Not recovered: audience-building and ad-pixel features. Retargeting audiences built from pixel page views will still miss blocked visitors.
  • Consent still applies. A server call does not cancel a visitor’s choice. See Does server-side tracking bypass consent? and Consent Mode v2 on WordPress.
  • Not a way around people. A visitor who installs a blocker is saying something about tracking. GDPR questions are covered in GDPR and WordPress analytics: the Danish picture.

Traculo’s part in this is narrow. With the WordPress plugin and a domain API key, events are forwarded from Traculo’s servers to the ad and analytics platforms you connect. That last hop does not pass through the visitor’s browser. Traculo’s own browser-side analytics script is a script like any other and can be matched by filter lists; I am not claiming it is exempt. Read the client-side vs server-side comparison and the no-GTM server-side guide for the trade-offs.

What to do in each situation

Decision table: situation and recommended action for a small gap, a noticeable gap with paid ads, a large gap from a technical audience, and consent-limited data.
A starting point, not a rule set. Thresholds are judgement calls, not published benchmarks.

In prose:

  • Gap under about 10% and mostly consumer traffic. Do nothing structural. Note the gap in reporting and move on. (The 10% is my rule of thumb, not a source.)
  • Gap noticeable and you run paid campaigns. Forward purchase and lead events server-side so ad platforms see conversions that pixels missed. Deduplicate against the browser event with a shared event ID; see Meta CAPI deduplication.
  • Gap large because of a technical audience. Accept that a client-side tool will undercount you. Use server logs for traffic volume, and your orders or CRM for conversions, as ground truth.
  • You need cleaner trends, not exact counts. Keep one tool consistently. Tool-to-tool comparisons in this area are noisy; trends are still readable.
  • You are shopping for a replacement. A different vendor changes how often you are matched by lists, not whether you are. See the GA4 alternatives roundup once published.

Frequently asked questions

What percentage of people use ad blockers?

It depends on the source. eyeo, which makes ad-filtering software, reports about 30% of the internet population in Q4 2025 and a country table from 15% (Japan) to 41% (Indonesia). Treat any single global figure as an estimate from an interested or secondary source.

Do ad blockers block Google Analytics 4?

Public filter lists contain rules for Google’s analytics and tag manager hosts and script paths, and Plausible reports GA missing most visitors in one tech-heavy test. Blocking is not universal: it depends on the blocker, lists, browser and site.

Can I recover blocked page views with server-side tracking?

Not the page views themselves, unless the page view is also collected server-side or from your logs. You can recover conversion events sent from the server.

Does first-party hosting of the script solve it?

Partly. It avoids domain rules for third-party hosts, but path rules such as /gtm.js match on any host, and visitors who block your own domain or all JavaScript still go uncounted.

Do ad blockers affect the numbers in Search Console?

Not directly. Search Console clicks are recorded on Google’s side rather than by a script on your page, which is why they make a useful sanity check, with the caveats above.

Related reading

To see which browsers and devices your visitors use before you decide how much this matters, open the breakdowns in Traculo’s website analytics. The free plan covers 10,000 events a month.

ShareLinkedInXEmail

About the author

Track what matters, without the guesswork

Traculo gives WordPress sites analytics, funnels and server-side event forwarding in one dashboard. Start on the free plan with 10,000 events a month.

Try Traculo free

Related articles

Title card: Do you need a cookie banner for WordPress analytics?
Privacy & ComplianceNovember 20, 2026

Do you need a cookie banner for WordPress analytics?

Whether WordPress analytics needs a cookie banner depends on what you store and your country's rules. CNIL, Denmark, Ireland, Germany compared.

Read more
Title card: GDPR and WordPress analytics: the Danish picture
Privacy & ComplianceNovember 18, 2026

GDPR and WordPress analytics: the Danish picture

What Datatilsynet, DIGST, the CNIL and the EU-US Data Privacy Framework say about WordPress analytics in Denmark, as of October 2026. Not legal advice.

Read more
Title card: Does server-side tracking bypass consent?
Privacy & ComplianceNovember 13, 2026

Does server-side tracking bypass consent?

Moving tracking to the server does not remove the need for a lawful basis or consent. What EDPB, CNIL, Datatilsynet and ad platforms say, plus a checklist.

Read more