How to Deploy Consent Mode for Google Ads

Google Ads Consent Mode (v2) Guide: GDPR Compliance, Performance and Measurement

Since 2024, advertising measurement has had to contend with a lasting reality: more users opting out of cookies, stricter regulatory requirements, and growing pressure to keep tracking reliable. Consent mode for Google Ads (v2) sits squarely in that tension: respecting user choice (and GDPR) whilst preserving measurement capability to steer conversions and performance.

This guide focuses on v2: what it is, Google's requirements, technical implementation (gtag and Google Tag Manager), key settings (ad_storage, analytics_storage and v2 signals), how it works with a CMP, and how to verify it in Google Ads. It deliberately sidesteps campaign creation and general Google Ads basics, so it stays practical for marketing, analytics and technical teams.

What This Guide Covers: Consent, Technical Implementation and the Impact on Conversion Tracking

In the rest of this article, you will find:

• Google's requirements and prerequisites (and what they really mean for your tags);
• the difference between basic and advanced implementation (tag loading, data sent, modelling);
• configuration examples for key signals (ad_storage, analytics_storage, ad_user_data, ad_personalization);
• a testing method: Tag Assistant, GTM Preview mode, network inspection (the gcs/gcd parameters);
• real-world impacts on observed conversions, attribution and bidding strategies.

Why Google Evolved This Mode: Privacy, Signal Loss and Ads Requirements

The shift is happening on two fronts: on one side, privacy expectations (GDPR, ePrivacy and DMA-related requirements in Europe); on the other, signal loss that weakens optimisation. When a user refuses certain purposes, Google distinguishes non-personalised ads (influenced by context, session activity and general location) from personalised ads (based, amongst other things, on prior browser activity), as explained across Google's privacy interfaces. The challenge is not "ads or no ads", but "which data, which storage, and what type of measurement".

At the same time, measurement still needs to be steerable. Google indicates that after an implementation has been in place for at least 7 full days, Google Ads and Google Analytics may have enough data to report an uplift (provided minimum data thresholds are met). In other words: the quality of your implementation directly determines how well Google can fill measurement gaps via modelling.

Google Consent Mode v2: Definition, Requirements, Prerequisites and Deadlines

Version 2 is an evolution of the framework that lets you send more granular consent signals to Google, especially for advertising use cases. One key point, repeated by Google Ads Help: this mechanism does not collect consent. It assumes you already have a compliant cookie banner or CMP (Consent Management Platform), and it is used to pass consent status to Google tags so they adapt their behaviour (cookie writing, data sending and measurement).

What Are Google's Requirements for v2?

The requirements boil down to three operational principles:

• Have a consent collection mechanism (a CMP or an in-house banner) before thinking about technical implementation.
• Transmit the status (refused/granted) to Google via the consent APIs so tags adjust automatically.
• Ensure your tags actually respect that status across all pages (governance, testing and trigger control).

In practice, this usually means setting a default state (often denied) and then updating it (often to granted) after user interaction. Google explicitly recommends this approach for an in-house banner in order to maintain conversion tracking and remarketing when the user accepts.

What Changes Between v1 and v2 for Ads: New Conditions for Personalisation and Measurement

v2 introduces additional advertising-focused signals that complement the historical storage parameters. Alongside analytics_storage and ad_storage, v2 notably adds ad_user_data and ad_personalization. The aim is to better distinguish:

• storage (cookies/identifiers) for advertising and analytics;
• use of user data for advertising;
• ad personalisation.

This granularity matters if you want tag behaviour to match user choice—and to avoid an all-or-nothing implementation that creates inconsistencies (or compliance risks).

Understanding the Deadline: Business Risks If You Are Not Ready

Google presented this evolution as mandatory from March 2024, notably in the context of the DMA in Europe. If your deployment is not ready, the most common risk is not that campaigns stop, but a drop in measurement quality: under-reported conversions, less stable attribution and more erratic learning for smart bidding.

From a compliance perspective, the opposite risk is also real: if marketing tags fire before valid consent, you expose yourself to GDPR non-compliance (particularly around lawful basis and governance of cookie writing/reading). The target scenario is a double win: properly gate execution and keep measurement coherent when the user accepts—whilst managing privacy-safe signals when they refuse (if you choose advanced mode).

How v2 Works: Signals, Pings and Modelling

The mechanism is straightforward: your Google tags receive a consent state (default, then updated) and adapt cookie writing and data sending accordingly. According to Google, the state is then preserved as the user navigates, to avoid inconsistent behaviour from one page to another.

With and Without Consent: Expected Tag Behaviour and Signal Transmission

Google describes two behaviours depending on the implementation you choose:

• If the user consents, cookies can be written and full measurement data can be sent.
• If the user refuses, behaviour depends on the mode (basic vs advanced). In advanced mode, cookieless signals (pings) can be sent whilst still respecting the refusal of storage.

Google notes that in advanced mode, these pings can include functional information (timestamp, user agent on web, referrer) and coarse information (limited granularity) such as the presence of ad click parameters in the URL (GCLID/DCLID) or boolean indicators of consent state.

Conversion Modelling: Principle, Prerequisites and Limits

Modelling aims to reduce measurement gaps when cookies cannot be set or read. Google indicates its products can use pings to model certain metrics, with one important constraint: some models only apply beyond a minimum collection threshold (Google states a threshold exists but does not disclose the value).

The practical implication: if your traffic is low, or if your implementation sends inconsistent signals (bad default/update ordering, tags blocked by a CMP in advanced mode, duplicates), you may not benefit from useful modelling. That is why technical QA and tag governance are essential.

What Remains Measurable Without Cookies: Realistic Expectations for Reporting

Without consent, expect more limited measurement: fewer directly observed conversions and greater reliance on modelling (if eligible). In advanced mode, you at least retain consent-state signals and privacy-safe pings. In basic mode, Google states nothing is transferred before interaction and, if the user refuses, no data is sent (including consent status, because tags remain blocked).

Basic vs Advanced: Choosing the Right Mode for Your Google Ads Set-Up

Google Ads Help distinguishes two implementation modes. The choice is not purely technical: it influences measurement continuity, modelling quality and your ability to diagnose what really happens when users refuse.

Logic Differences: Minimal Adjustment vs Enriched Signals

• Basic mode: tags do not load until the user interacts with the banner. No data is sent to Google before interaction. If the user refuses, nothing is sent.
• Advanced mode: tags load on page open, set default states (often denied), send cookieless pings until consent is granted, and then adjust after the user's choice.

Google also indicates that modelling in advanced mode relies on an advertiser-specific model, whereas basic mode relies on a more general model.

Impact on Conversion Tracking and Attribution: Where the Gap Shows Up

The differences are most visible in three areas:

• Conversions: fewer observed conversions when refusals are common, with potential partial compensation via modelling (more likely in advanced mode).
• Attribution: more uncertainty when identifiers and cookies are unavailable.
• Diagnosis: in basic mode, tag silence on refusal makes it harder to understand signal loss.

Decision Criteria: Volume, Geography and Legal Constraints

To choose, combine three factors:

• Traffic and conversion volume: higher volume increases the odds of meeting modelling thresholds.
• Geography: in Europe, consent requirements are central. Your default configuration and banner must reflect local reality.
• Internal legal constraints: some organisations prefer an ultra-conservative approach (basic) at the cost of weaker measurement; others choose advanced, subject to DPO validation and documented mechanics.

To align paid performance with organic visibility more broadly, you can also compare the logic of SEO vs SEA to keep KPIs, measurement and investment consistent.

Consent Settings: ad_storage, analytics_storage and Other v2 Signals

These settings control what can be stored and how Google products behave. A strong practice is to separate purposes clearly—analytics, advertising, personalisation—which makes it easier to align CMP choices with tag behaviour.

ad_storage: What You Allow (or Not) for Advertising

ad_storage controls advertising-related storage. Simplified: if ad_storage is set to denied, you must not allow advertising cookies to be written/read. If the user accepts, switch to granted to re-enable compatible advertising features (measurement and, depending on your set-up, remarketing).

analytics_storage: Align Measurement With Your Analytics Set-Up

analytics_storage governs analytics storage (for example, GA4). A common pitfall: a banner may allow analytics but refuse marketing. In that case, you must reflect the choice by separating analytics_storage from ad_storage rather than switching everything together.

v2 Signals for Ads: ad_user_data and ad_personalization, Impacts and Prerequisites

v2 adds signals dedicated to advertising use cases:

• ad_user_data: permission to use certain user data for advertising.
• ad_personalization: permission to personalise ads.

In practice, these signals strengthen alignment between user choice and the type of ads served (personalised vs non-personalised), as described in Google's interfaces. They require a CMP capable of expressing sufficient granularity (and an implementation that does not flatten everything into a single on/off button).

Getting ad_storage and analytics_storage Right: Default States, Updates and Execution Order

The recommended pattern is:

1. Set a default state (often denied) as early as possible, ideally before any tags can fire.
2. Update immediately after the user's decision (with granular accept/refuse where relevant).
3. Control ordering: the default must come before the update, otherwise you create non-deterministic behaviour (tags firing before state is known).

Typical example (based on GTM-documented recommendations and common practice):

gtag('consent', 'default', {
ad_storage: 'denied',
analytics_storage: 'denied',
ad_user_data: 'denied',
ad_personalization: 'denied'
});

Then, after acceptance:

gtag('consent', 'update', {
ad_storage: 'granted',
analytics_storage: 'granted',
ad_user_data: 'granted',
ad_personalization: 'granted'
});

Technical Implementation: Deploying and Enabling Consent Mode Correctly

Deployment depends on your stack (website, SPA, CMS), your operating model (marketing autonomy vs dev team), and tag governance. Google notes that implementation varies by infrastructure and can rely on a partner CMP, an in-house banner or (for apps) specific SDKs.

Choose Your Architecture: gtag, Google Tag Manager or a Hybrid Approach

Three architectures are common:

• gtag only: fast, but requires strict discipline to avoid scattered hard-coded scripts.
• Google Tag Manager: central orchestration, versioning and QA. If you are starting out or need to harden your stack, GTM remains the control centre for triggers.
• Hybrid: gtag for the consent layer + GTM to deploy tags. It works, but increases the risk of duplication if responsibilities are unclear.

If you use GTM, a good starting point is to understand the container/tag/trigger/variable model and stabilise your events via the dataLayer, especially on B2B journeys (forms, conversion steps, SPAs).

Default Consent by Region: Best Practice and Common Mistakes

Best practice is to set a default state that matches your region (often denied in Europe) and push it very early (GTM "Consent Initialization" trigger). Common mistakes include:

• Setting the state too late (after some tags have already loaded);
• Letting a CMP block tags when you are targeting advanced mode (Google recommends checking this and unblocking where needed);
• Double implementation (hard-coded tag plus GTM), which can double pageviews, events or conversions and make diagnosis unreliable.

Updating the User's Choice: Timing, SPAs and Transitional Pages

The update should happen immediately after the user's choice in the banner/CMP. On SPAs, the main challenge is persisting state across route changes: ensure consent state is stored (by the CMP) and correctly applied to subsequent events, without falling back to the default denied on every internal navigation.

On transitional pages (checkout, subdomains, redirects), consistency is the priority: a choice made on one page must apply across the entire journey. This also applies to non-Google tags fired via GTM, which should follow the same consent rules.

Enabling URL Passthrough: Use Cases, Benefits and Limits

URL passthrough mainly helps preserve click information in the URL when cookies cannot play their usual role. In practice, it can support continuity of measurement in certain redirect-heavy journeys, but it is not a silver bullet: you still need to respect consent and avoid cross-domain inconsistencies.

High-Risk Cases: Redirects, Cross-Domain and Self-Referrals

Three recurring risks are:

• Redirects that drop parameters or cause tags to fire in the wrong order.
• Cross-domain complexity (multiple domains, subdomains, third parties) that weakens session continuity and attribution.
• Self-referrals in analytics tools, where redirect chains rewrite source/medium and degrade reporting.

CMPs and Cookie Banners: Sync Consent, Proof and Google Tags

The key point (from Google's perspective) is that the framework relies on your CMP/banner to capture user choice. Your CMP owns the consent UX; your tags must execute that choice without leakage (premature firing, cookies set too early).

How Does Consent Mode Interact With a CMP?

In a standard set-up, the CMP:

• displays the banner and captures the choice;
• exposes the choice (via callbacks, variables or the dataLayer);
• triggers the state update (via gtag or GTM).

Google lists CMPs with integration documentation. In all cases, the same watch-out applies: many banners block Google tags by default before consent. If you are targeting advanced mode, you must check and adjust that blocking.

What the CMP Must Control: Purposes, Granularity and Consent Evidence

To be usable and defensible, your set-up should cover:

• Distinct purposes (necessary, measurement, marketing) at a minimum.
• Sufficient granularity (beyond "accept all / reject all"), aligned with your tags. Google interfaces also reference the idea of a "More options" path for privacy settings.
• Evidence and traceability (who consented, to what, when, and how users can change their choice).

Asynchronous Loading: Preventing Premature Tag Firing

A common non-compliance scenario comes from poorly orchestrated asynchronous loading: a marketing tag fires before the "default denied" state is applied, or before the CMP has exposed the decision. Mitigation requires:

• a "Consent Initialization" trigger for the default;
• tag-level consent settings (in GTM, advanced consent configuration) to enforce granted prerequisites where needed;
• regular audits of third-party tags, often responsible for "ghost" firing.

Withdrawal and Preference Changes: Keeping Long-Term Consistency

Compliance does not end with the first choice. You need:

• an easy way for users to change preferences;
• effective updates on the tag side (not only in the CMP interface);
• consistent behaviour across all pages and environments (staging and production).

GDPR Compliance: Responsibilities, Documentation and Key Watch-Outs in France

In France, GDPR and tracker-related rules impose a strict framework: clear information, free and specific consent, and the ability to withdraw consent. Consent must control firing and storage—not the other way around.

What Consent Mode Helps With (and What It Does Not) for GDPR Compliance

This framework provides a technical capability: making Google tags respect user choice by adjusting cookie writing and data sending. However, it does not provide:

• a compliant banner;
• consent evidence (that is the CMP's and your governance's role);
• a turnkey legal validation (you still need to define purposes and lawful basis).

Valid Consent: Information, Freedom of Choice, Granularity and Retention

To avoid "fake implementations", check that:

• users understand purposes (measurement vs advertising vs personalisation);
• rejecting is as easy as accepting;
• preferences can be adjusted at any time;
• states sent to tags reflect those preferences exactly.

From a data minimisation standpoint, a more privacy-respecting approach also builds trust. For context, according to Squid Impact (2025), anonymisation increases trust by +34% and 81% of companies prioritise anonymisation for trustworthy AI. These figures are not specific to Google Ads, but they illustrate the wider trend: transparency and minimisation are becoming structural advantages.

Auditing and Traceability: What to Keep to Secure Your Practices

Strong traceability is both a GDPR requirement and a marketing reliability issue. Keep:

• published GTM versions (what changed, why and when);
• screenshots/exports from Tag Assistant and network inspection showing default then update;
• a tag-and-cookie map by purpose, including firing conditions;
• a measurement plan (events, parameters and de-duplication rules).

Impact on Google Ads Conversion Tracking: Interpretation, Quality and Performance

The first visible effect of a correct deployment is not always an increase in conversions, but more coherent measurement aligned with consent—plus fewer tracking breaks. That changes how you interpret reports, especially during the transition period.

What Changes in Tracking: Tags, Consent and Signal Loss

When consent is refused, directly observed measurement drops mechanically (less storage, fewer identifiers, sometimes no sending at all in basic mode). This creates gaps that compound other common causes (ad blockers, redirects, processing delays, attribution models and tag duplication).

In a context where digital performance heavily depends on Google (89.9% global market share and around 8.5 billion searches per day, according to Webnyxt, 2026), protecting measurement becomes an operating capability—not just a technical toggle. To follow the trends shaping visibility and measurement, you can review our SEO statistics and GEO statistics.

Enhanced Conversions: Compatibility, Prerequisites and De-duplication

Whatever your set-up, de-duplication is a major prerequisite. A common mistake is double-counting a conversion (hard-coded tag plus GTM) or firing on a fragile signal (e.g. simply loading a page) rather than a true success event (ideally pushed into the dataLayer). On B2B journeys, separate macro-conversions (demo request, quote request) from micro-actions (CTA click, form open) to avoid over-optimising weak signals.

Effects on Bidding Strategies: Learning, Stability and Control

When observed conversions fall, bidding strategies can enter a more volatile learning phase (less signal, more variance). The goal is to preserve sufficiently continuous measurement to keep trends usable. In advanced mode, pings and potentially richer modelling can reduce the harshness of the break—but it depends on your volume and implementation quality.

Before/After Measurement: Comparing Without Misreading the Data

Avoid a simplistic 48-hour before/after comparison. Plan for:

• a meaningful observation window (at least one to two weeks, and remember Google's reference to 7 full days for certain insights);
• annotating changes (GTM version, CMP changes, trigger changes);
• segmented analysis: region, device (mobile accounts for 60% of global web traffic, according to Webnyxt, 2026), landing pages and campaigns.

How to Verify Everything Works in Google Ads (Tests, Tools and Checklist)

Checking "in Google Ads" alone is not enough. Effective validation combines debugging tools, technical inspection and business logic (the conversion fires at the right moment).

How Do You Test Whether Consent Mode Is Working Properly?

Work in three steps:

1. Functional test: reject, accept, partially accept (analytics yes / ads no), withdraw/change choice, then repeat.
2. Technical test: verify the default → update order and final state for each signal.
3. Business test: complete a real conversion (form, purchase) and verify it is received (real-time/debug first, then reports).

Check the Signals Being Sent: Network Inspection and Diagnosis

Using Tag Assistant and DevTools, inspect network requests and check consent-related parameters (often documented via gcs and gcd). The goal is to confirm the transmitted state matches the user's choice—and changes at the right time.

Validate Firing: GTM Preview Mode and Control Points

In Google Tag Manager, use Preview mode. The "Consent" tab lets you check the default value, updates and current state. Also verify:

• no marketing tags fire before the default;
• tags requiring ad_storage do not fire when it is denied;
• your dataLayer events have stable values (no empty or shifting parameters).

For a deeper method, you can read our guide on how to test a GTM implementation (Preview logic, tags fired and variable checks).

Check Within Google Ads: Indicators, Alerts and Conversion Consistency

In Google Ads, focus on consistency: a conversion should not disappear entirely (if the user accepts), nor be duplicated. Monitor tag alerts, sudden volume gaps, and conversion stability by campaign and device. Remember that changes can come from consent—but also from site updates (SPA routing, redirects), a rushed GTM publish, or competing tags.

Common Issues: Consent Not Updating, Duplicates and Competing Tags

• Consent not updating: state remains denied despite acceptance (often a CMP callback or execution-order issue).
• Duplicates: the same conversion fires from two sources (hard-coded plus GTM), or two different Ads tags.
• Competing tags: a CMS plugin, partner or third-party tool injects tags outside GTM and ignores consent logic.

Maintenance and Governance: Avoiding Tracking Debt After Implementation

Tracking debt appears when tags accumulate, consent rules diverge, and versions multiply without QA. And a slow website is expensive: Google stated in 2025 that 40% to 53% of users leave a site if it loads too slowly. Every unnecessary script is therefore a performance risk—and a conversion risk.

Versioning and QA: Securing Banner, Tag and Site Changes

Adopt a simple routine:

• staging then production (separate environments);
• GTM publishing with a clear description (scope and impacted tags);
• a standardised test plan (reject/accept, conversion, network inspection).

Ongoing Monitoring: Catching Regressions Before They Cost Conversions

Continuously monitor conversion volumes, tag error rates, load time and cross-tool consistency. When a gap appears, look for a technical cause first (GTM version, CMP, redirect) before concluding it is a marketing fluctuation.

Frictionless KPI Management: How Incremys Helps You Measure Impact and ROI

Centralise Your Analysis: SEO/GEO Tracking, Dashboards and Automated Reporting

Once measurement is stable, the next challenge is management: comparing periods without misreading them, linking effort (content, pages, campaigns) to outcomes, and making metrics actionable. Incremys can support this stage through its Incremys performance reporting module, which helps centralise KPIs and automate dashboards. The goal is not to replace your tracking tools, but to enable a more structured readout and more consistent measurement of impact and ROI.

To strengthen overall performance beyond tracking, you can also structure your SEO & GEO content production strategy and build authority through Google netlinking, two levers that complement Ads campaigns.

If your goals also include local visibility, Google Maps SEO can become a key channel for capturing high-intent demand, with measurement aligned to the same consent rules.

Finally, to anticipate how changes (consent, seasonality, publishing and competition) may affect performance, you can use our predictive AI to prioritise actions and investment more effectively.

FAQ: Common Questions About v2, Consent and Google Ads

What Is v2 and Why Does Google Require It?

v2 is an evolution of consent signals, with stronger granularity for advertising use cases (ad_user_data, ad_personalization) alongside ad_storage and analytics_storage. Google requires it to better align personalisation, measurement and user choice in a tighter regulatory context in Europe.

What Deadline Should You Follow, and What Actions Should You Prioritise?

Google tied this change to a requirement starting in March 2024. Prioritise: (1) a compliant CMP/banner, (2) correctly ordered default and update states, (3) tag firing control, and (4) a test plan and documentation.

What Is the Difference Between Basic and Advanced?

In basic mode, tags do not load before interaction and if the user refuses, nothing is sent. In advanced mode, tags load with a default state (often denied) and may send cookieless pings until consent is granted, which generally enables richer modelling on Google's side.

How Do You Configure ad_storage and analytics_storage?

Set a default state (often denied) for both ad_storage and analytics_storage, then update based on the user's choice. Example: set default to denied on load, then update to granted only for the purposes that were accepted.

What Is the Best Technical Implementation for Your Stack?

For a site with frequent marketing changes, GTM makes governance easier (versions, QA and consent rules). For a dev-controlled stack, gtag may be sufficient. A hybrid can work, but requires strict discipline to avoid duplication and inconsistent execution order.

How Do You Work With a CMP Without Firing Too Early?

Apply the default state very early (before any tags), then trigger the update via a reliable CMP-emitted event. In GTM, reinforce with tag-level consent settings to require granted where necessary.

How Do You Perform an Effective Validation to Confirm Collection and Measurement?

Combine GTM Preview (Consent tab), Tag Assistant, network inspection (gcs/gcd), and a real conversion test. Verify the default then update order, and confirm there are no competing tags.

What Is the Impact on Conversion Tracking and Campaign Performance?

Expect fewer observed conversions when refusals are common, especially in basic mode. In advanced mode, cookieless signals can support modelling (subject to thresholds and implementation quality), which can preserve some reporting clarity and bidding stability.