Consent Mode v2 is Google’s framework for adjusting tag behavior based on user consent choices. Mandated for EEA/UK traffic since March 2024 and enforced with automated account restrictions since July 2025, it affects every Google Ads and GA4 implementation that serves European users. Getting it wrong silently corrupts data; getting it right requires understanding the parameter model, implementation sequencing, and server-side propagation.
This hub connects the concepts involved in a complete Consent Mode v2 implementation.
Core Concepts
Consent Mode v2 Parameter Architecture — The four consent parameters, how upstream browser controls (ad_storage, analytics_storage) differ from downstream server instructions (ad_user_data, ad_personalization), and the EU Digital Markets Act legal mandate that forced the expansion from two to four parameters.
Consent Mode Basic vs Advanced — How each mode handles declined consent. Basic mode blocks all tags; Advanced mode sends cookieless pings. The traffic thresholds for advertiser-specific conversion modeling (700+ ad clicks for Google Ads, 1,000+ daily events for GA4) and when Advanced mode’s modeling advantage actually materializes.
Implementation
Consent Mode Implementation Mechanics — The three components that must work together: default consent states (fired in Consent Initialization trigger), CMP integration (all four v2 parameters in the update call), and the wait_for_update parameter that prevents the async CMP race condition. Includes region-specific defaults and non-GTM implementations.
Consent Mode Server-Side GTM Propagation — How consent signals travel from web container to server container via gcs and gcd HTTP parameters. Google tags handle this automatically; non-Google vendor tags (Meta CAPI, TikTok Events API, LinkedIn CAPI) require manual consent enforcement through trigger conditions.
Debugging
Consent Mode Debugging Network Parameters — How to decode the gcs and gcd parameters in browser network requests. The verification workflow for four consent states (initial load, accept, deny, partial), what each letter code means, and the tools that simplify inspection (Consent Mode Inspector extension, Tag Assistant).
Failure Modes
Consent Mode Common Implementation Failures — The ten most frequent implementation mistakes, ordered by prevalence and damage: missing defaults, v1-only parameters, broken server-side propagation, missing region configuration, race conditions, unwired banners, duplicate tags, case sensitivity, Enhanced Conversions without ad_user_data, and insufficient testing across consent states.
US Considerations
Consent Mode US Privacy Requirements — Why US-only sites increasingly need Consent Mode: Google’s product requirements for Enhanced Conversions and remarketing, the expanding state privacy landscape (20 states with comprehensive laws), and the recommended region-specific configuration for CCPA opt-out models.
Related Concepts
GA4 BigQuery Number Discrepancies — Consent Mode behavioral modeling is typically the largest source of variance between GA4 interface numbers and BigQuery exports. Users who reject consent are modeled in the GA4 interface but absent from BigQuery.
Article Source
These notes were decomposed from Consent Mode v2: implementation, enforcement, and the ten mistakes breaking your data, which provides the complete narrative covering the July 2025 enforcement, implementation details, all ten failure modes, debugging procedures, US requirements, and the December 2025 enforcement changes.