An analytics category can cover simple page counts, individual browsing recordings, and advertising attribution. Review the enabled purpose of each integration before deciding what the category's permission allows.
A tool's name or its position in a cookie declaration doesn't establish anonymity, an exemption, or consent for every feature the vendor offers.
Compare three uses labeled analytics
Change the enabled use while keeping an analytics description. Inspect when the example’s description stops matching its purpose.
- Banner description
- Site usage statistics
- Enabled use
- Aggregate page counts to improve the site
- Additional evidence
- Retention, aggregation, and provider-use settings still need review
This is a candidate for reviewing a statistical exemption. The word analytics doesn't establish eligibility.
Next check: Review the actual configuration and jurisdiction-specific conditions.
Invented observations for teaching. Manual-debugger details illustrate evidence beyond CookieTrail’s current standard capture. Colors describe this example’s evidence or stated policy, not legal compliance.
List the enabled purposes
Aggregate counts answer questions such as which pages are popular. Individual recordings let a team inspect a person's browsing session. Attribution connects activity to an advertising outcome. These uses need separate descriptions and permission decisions even when a team groups them under measurement.
- What information is collected before any choice?
- Can the team revisit an individual session or identify a returning visitor?
- Does the integration connect events to advertising clicks, audiences, or conversions?
- Does the provider reuse the information or combine it with other data?
- Which optional features are enabled on this deployment?
Separate product capability from observed use. A vendor's documentation can tell you what is possible; configuration and browser evidence show what this site has enabled.
Check the boundary of a statistics exception
The UK's statistical exception concerns aggregate results used to improve the service. ICO guidance excludes individual visitor recordings and advertising measurement from that exception. CNIL's French criteria also restrict data reuse and cross-site tracking.
- ICO exception boundaries Examples separating statistical measurement from individual tracking.
- CNIL measurement exemption Purpose and reuse restrictions for French audience measurement.
Treat a site's claim of anonymous analytics as a statement to verify. Masking a form field does not establish that an entire dataset is anonymous. A label such as non-identifiable also leaves questions about persistent identifiers, linkage, retention, and provider use.
Test an analytics-only choice
- Start a fresh session in a verified visitor region and record the categories and their explanations.
- Allow analytics while leaving marketing disabled. Confirm that the saved categories match that choice.
- Inspect the enabled integrations and their events. Distinguish script downloads from collection and identify the evidence for any advertising use.
- Compare with separate reject-all and accept-all sessions at the same URL. Retest after any category or feature change.
This selective test can reveal a mismatch that accept-all and reject-all do not exercise. CookieTrail's standard audit currently records first load, accept all, and reject all; an analytics-only test needs a separate manual or custom capture.
Describe a purpose mismatch precisely
If a site describes an integration as aggregate statistics but manual evidence shows advertising attribution enabled with marketing denied, preserve both facts. That is more useful than declaring that a particular brand always belongs in one category.
For a Meta Pixel example, check the deployment's actual events and features, its description, and the relevant permissions. A single request to a recognizable host does not reveal all downstream processing.
Then compare the banner's refusal promise with the same behavior.
A study of category mismatches needs reviewed declarations and selective-choice evidence. Count eligible sites where the relevant integration and category were established. Keep unresolved classifications separate, and explain which purposes the evidence can actually identify.