The State of Web Tracking 2026
A data-driven look at which analytics, tag-management, consent, and CDP tools actually run on the web — measured from public HTTP Archive crawl data, not surveys.
Enter your email — the PDF lands in your inbox and you're on The Tracking Wire.
Weekly. Free. One-click unsubscribe. Opens Substack in a new tab — you won't lose your place here.
This is the flagship report and the lead magnet behind the newsletter. The methodology is the point: findings come from the HTTP Archive public dataset on BigQuery — a monthly crawl of millions of sites — cross-referenced with open technology fingerprints. No vendor surveys, no self-reported numbers.
What the report covers
- Analytics market share — GA4 vs Adobe vs the privacy-first field (Plausible, Umami, Matomo), by site rank tier.
- Tag-manager adoption — GTM, Tealium, Adobe Launch/Tags, and the rise of server-side containers.
- Consent & CMP — TCF v2.2, GPP, and Consent Mode v2 prevalence, and how often tags actually respect them.
- CDP & warehouse-native footprint — Segment, RudderStack, and the shift toward BigQuery/Snowflake-native stacks.
- The performance tax — tags-per-page distributions correlated with Core Web Vitals.
How it’s built
The pipeline is reproducible: parameterized BigQuery queries against HTTP Archive, post-processed with Polars, and rendered into both the web report and a gated PDF. The same fingerprint library powers the free Pixel & Tag Scanner — the report is the aggregate, the scanner is the single-site view.
Data drop pending — methodology below
The queries are written and the pipeline runs end-to-end, but no findings are published here until the first full BigQuery run against a real crawl completes. We will not print a number we did not measure.
What the first data drop will be measured from:
- Source: the HTTP Archive public BigQuery dataset — a monthly crawl of millions of sites recording detected technologies (Wappalyzer-based), full request data, and Core Web Vitals via CrUX. Public data intended for research; every query we run will be published.
- Scope: root pages of ranked origins, one crawl month per snapshot, with results broken down by CrUX site-rank bucket (top 1k / 10k / 100k / 1M).
- Measures: adoption by category (analytics, tag management, consent/CMP, CDP, experimentation); market share within categories; trends across crawl months; per-page tracking-request share computed with our open vendor fingerprint library (the same rules behind the Pixel & Tag Scanner); third-party domain and cookie-setting prevalence; and tracking weight correlated with CWV.
- Cost/repeatability discipline: date-partitioned, column-pruned queries prototyped on HTTP Archive’s sample tables, dry-run before every paid scan, with small materialized extracts as the only build input — so every refresh is cheap and repeatable when the next monthly crawl lands.
Subscribe on the home page to get the full PDF when the first data drop ships.