Consent Mode adjusts tag behaviour according to the visitor’s consent. Configuring it is one thing; making sure it works is another. Three simple tests, which any team can run, show whether users’ choices are respected and whether measurement is not lost unnecessarily.
The essentials in a definition
Auditing Consent Mode in practice means verifying, with simple tests, that consent signals reach tags correctly and tags respect the user’s choice. Three tests suffice: default state before the choice, behaviour after accepting and behaviour after refusing. The developer instructions, GA4 documentation and GDPR define the requirements.
What Consent Mode does
Instead of blocking everything when a visitor refuses, it lets tags send limited signals, without cookies or identifiers, so platforms can do modelling. When the visitor accepts, tags run normally.
This requires the cookie banner to communicate the consent state correctly to the tags before they fire.
Test 1: default behaviour
Open the site in a clean browser without accepting anything. Check, with debugging tools, whether the initial consent state is denied for the relevant purposes and whether there are no advertising cookies before the choice.
If there are cookies or requests to ad platforms before the visitor decides, there is a configuration problem.
Three Consent Mode tests
- Default state: nothing before the choice
- Accept: full tags firing
- Refuse: only permitted signals
- Repeat on devices and after changes
Tests 2 and 3: accept and refuse
Accept and confirm that the state changes to granted and tags fire with all parameters. Then refuse and confirm it remains denied and only permitted signals are sent.
Repeat on different devices and browsers. Also test changing the choice later: what happens if the user withdraws consent?
Consent only protects if it is respected before the first tag fires.
After the tests
Document the results and fix what fails. Track the platforms’ consent diagnostics and the percentage of modelled conversions. Repeat the tests whenever you change the cookie banner, the site or the tag manager.
Validate decisions with whoever handles data protection. Correct technical configuration is necessary but does not replace legal analysis.
Decision table
| Test | How to do it | Expected result |
|---|---|---|
| 1. Before the choice | Open the site in an incognito window and inspect requests | No storage or sending of disallowed data |
| 2. After accepting | Accept all and browse | Tags fire with the right signals |
| 3. After refusing | Refuse and browse | No identifiers; only permitted signals |
| 4. Changing the choice | Change the preference | Tags respect the new state |
Common mistakes to avoid
- Accepting that “the banner is there”. A banner not integrated with tags is useless.
- Testing only once. Repeat after changes to the site and tag manager.
- Forgetting mobile. Behaviour may differ in apps and mobile browsers.
How to measure the result
Keep the test evidence (date, version, results) and track the consent rate and the proportion of modelled conversions after each change.
Limits and exceptions
Three tests do not replace a full privacy audit or legal advice. Implementations vary (consent management platforms, themes, plugins) and rules evolve. Technical tests show whether the system works as configured, not whether the configuration is suited to the law. Combine technical verification with legal review and update it whenever you change the site or suppliers.
Illustrative example
Illustrative example: a team discovers, by testing, that a remarketing tag fires before the user chooses. It fixes the load order and signal configuration, repeats the three tests and documents the results with screenshots. In the diagnostics report, consent rates now reflect reality.
Frequently asked questions
Is Consent Mode mandatory?
For certain advertising and measurement uses in the EU, it is recommended practice or a platform requirement. Check your case.
Who runs the tests?
Whoever manages the tags, reviewed by whoever handles data protection.
How often?
Whenever you change the banner, the site or the tags, and at least quarterly.
Who should do the audit?
Whoever manages measurement, with legal support when necessary; repeat after major changes.
Which tools should I use?
Browser developer tools, the tag manager’s preview mode and the platform’s diagnostic reports.
Three simple tests (before the choice, after accepting and after refusing), repeated after every change and documented, verify that Consent Mode respects the user’s choice.
Sources and recommended reading
- Analytics Help: consent mode
- Google for Developers: set up consent mode
- General Data Protection Regulation (GDPR)
Last editorial review: October 2026. Informational content; always confirm current conditions in the official sources before deciding.