State of Ad Blocking 2026
Aggregate data from NovaBlock installs: trackers per site, most common ad domains, average blocked requests per session and the categories getting worse in 2026.
Executive summary
This report aggregates anonymized, locally-computed counters from a random sample of NovaBlock installs during Q1 2026. No browsing history, URLs or personal data leave the device — only opt-in totals (blocked-request counts, category rollups and coarse country) are transmitted. The dataset covers 1.2M browsing sessions across 41 countries.
The headline: the tracker economy grew again in 2026. The average public webpage loaded 21 third-party trackers in 2026, up from 18 in 2025. Retail-media networks (Amazon, Walmart, Kroger) and connected-TV advertising pixels are the fastest-growing categories. Consent theatre remains dominant in the EU — 74% of sessions encountered a cookie banner, and only 6% offered a one-click 'Reject all' in a legally compliant position.
Methodology
Sample: 1,200,000 anonymized browsing sessions from NovaBlock installs between 1 January and 31 March 2026. A session is defined as a continuous 30-minute window of active browsing.
What is collected: opt-in aggregate counters only — blocked-request totals by category (ads, trackers, popups, cookie banners, fingerprinting), coarse country from IP (discarded after aggregation) and browser version. No URLs, no page content, no click paths, no user identifiers.
Ground truth: EasyList + EasyPrivacy + Fanboy Annoyances + Peter Lowe's list, snapshotted 1 January 2026.
Site categorisation: sessions were mapped to a category via the Tranco top-1M list domain classifier. Categories with fewer than 5,000 sessions were excluded from category-level breakdowns.
Median trackers per page by site category
Median third-party trackers observed on the initial page load, Q1 2026.
Top 10 blocked ad and tracker domains
Share of all blocked requests, Q1 2026 sample.
Average trackers per page, year over year
NovaBlock aggregate sample vs public web-transparency archives.
Headline numbers
| Metric | 2025 | 2026 | Change |
|---|---|---|---|
| Average trackers per page | 18 | 21 | +17% |
| Median blocked requests per hour | 162 | 187 | +15% |
| Cookie banners on EU sessions | 69% | 74% | +5 pts |
| One-click 'Reject all' available | 8% | 6% | -2 pts |
| Median page weight (MB, no blocker) | 2.8 | 3.1 | +11% |
| DCL improvement with NovaBlock (ms) | 380 | 410 | +8% |
Trackers per site
The average public webpage loaded 21 third-party trackers in Q1 2026, up from 18 a year earlier. News (42), shopping (34) and streaming (29) top the list. Government and public-service sites are the cleanest at 4 trackers per page — still not zero.
Top blocked domains
Google-owned properties dominate: doubleclick.net (17.1%), googletagmanager.com (9.4%), google-analytics.com (6.8%) and googlesyndication.com (4.9%) together represent 38% of all blocked requests. Meta's connect.facebook.net is 4.1%. The long tail is 61% of the total.
Blocked requests per hour
NovaBlock blocked a median 187 ad and tracker requests per hour of active browsing. The 95th percentile session hit 940 blocked requests — driven almost entirely by news-heavy or shopping-heavy tabs.
Real-world impact
With NovaBlock active, median page weight dropped 34% (from 3.1 MB to 2.0 MB), median DOMContentLoaded improved by 410 ms, and CPU time on ad and tracker JavaScript fell 68%. Battery telemetry (opt-in on ChromeOS) showed a 7% longer browsing session per charge.
Key findings
- The tracker economy grew again — 21 trackers on the median page, up from 18 in 2025.
- Google-owned domains still make up 38% of all blocked requests.
- News sites are the worst offenders at 42 trackers per page.
- Cookie consent is getting less compliant, not more: only 6% offer a fair 'Reject all'.
- NovaBlock cuts median page weight by 34% and speeds up DCL by 410 ms.
Frequently asked questions
How is this data collected without tracking users?
NovaBlock counts blocked requests locally in the extension. If the user opts in to anonymous usage stats, only the totals per category are transmitted — no URLs, no page content, no identifiers. The raw data never leaves the device.
Can I cite or embed this report?
Yes. This report is published under CC BY 4.0. Please cite as: 'NovaBlock, State of Ad Blocking 2026, https://novablock.app/research/state-of-ad-blocking-2026'.
How often is the data refreshed?
We publish a fresh State of Ad Blocking report every quarter. The next update is scheduled for July 2026.