Controlled-visit evidence · 5 canonical sites
What shopping sites loaded
Descriptive results from one newest eligible passive visit per site, not a ranking, privacy grade, causal claim or representative sample of all shopping sites.
Newest included observation:
Observed medians
Median across this curated sample.
Medians summarize the included site-level observations. They do not describe every visit to these sites or the wider category.
- Third-party requests
- 253
- Third-party tracking-service requests
- 50
- Third-party cookies
- 35 of 5 sites had complete cookie evidence
Included evidence
5 site-level observations
Sort any column. Open a profile for the currently retained, versioned report history.
5 sites
| amazon.comamazon.com contacted 1 cross-site host that the service catalog did not classify as tracking.1 report · Open latest evidence | 2 | 0 | 0 | Jul 27, 2026Brave-list blocking comparison · desktop |
|---|---|---|---|---|
| bestbuy.combestbuy.com recorded 203 fewer third-party requests in the visit configured for Brave-list blocking.1 report · Open latest evidence | 252 | 122 | 26 | Jul 27, 2026Brave-list blocking comparison · desktop |
| homedepot.comhomedepot.com sent data in personal-identifier fields to Meta Pixel.1 report · Open latest evidence | 307 | 53 | 14 | Jul 27, 2026Brave-list blocking comparison · desktop |
| target.comtarget.com recorded 80 fewer third-party requests in the visit configured for Brave-list blocking.1 report · Open latest evidence | 253 | 50 | 3 | Jul 27, 2026Brave-list blocking comparison · desktop |
| walmart.comwalmart.com recorded 12 fewer third-party requests in the visit configured for Brave-list blocking.1 report · Open latest evidence | 371 | 2 | 2 | Jul 27, 2026Brave-list blocking comparison · desktop |
How to read this page
Eligibility and limits
- Each row is the newest successful, request-complete, uncapped passive lead visit for one canonical site.
- Every row here was measured under one cohort (shields-request-context-v2-adblock-rust-0.13.2-request-method-v1-playwright-1.61.1, schema v1, producer unrecorded, GPC not requested), which is this page's denominator. Another category can publish a different cohort, so these medians are comparable within this page and not against another category's.
Full measurement identity
shields-request-context-v2-adblock-rust-0.13.2-request-method-v1-playwright-1.61.1, schema v1, producer unrecorded, GPC not requested, legacy catalog-metadata hash c015b2fb2d86a8aa1e015c740cb967f433f4c6301c0f4a010592f41f30945593, ServiceRole taxonomy service-role-taxonomy-v1 digest dfccf71d4119c154e71bf7908dd2914557e8fc981951941594b16b00b712ed67, metric contract metric-contract-v1 digest 19bd5b62d1530480481080a1f208d2664e3769ac16300560a1f3df5894eb9680
- Failed loads, incomplete recordings and accept/reject consent-interaction arms are excluded.
- Third-party tracking-service request counts are lower bounds derived from retained request rows whose recorded host matched a reviewed service-catalog suffix and from the recorded ServiceRole taxonomy.
- Category membership is editorial. The corpus is curated and is not a random or representative web sample.
- A controlled visit is one observation. Ads, experiments, caching, location and bot detection can change what loads.
- Blocking pairs simulate Brave's ad-block engine and default lists in the scanner; they are not live Brave-browser visits or counts of individually blocked requests.