Article · June 17, 2026

How does Consent Mode v2 affect conversion measurement?

How Consent Mode v2 affects conversion measurement, what you can still track when users refuse, and what data remains usable.

consent.vn Editorial6 min read

Quick answer

Consent Mode v2 does not “wipe out” measurement data, but it changes how conversions are collected and inferred. When users refuse, the system can use modeling to estimate the missing parts, while personalized/ads data is usually constrained by the consent status.

How does Consent Mode v2 affect conversion measurement?

Consent Mode v2 directly impacts the completeness of conversion reports, attribution, and remarketing. In short: if users consent, you can measure more; if they refuse, you can still receive limited signals and let the platform estimate via modeling, but results will no longer be as complete as before.

In practice in Vietnam, this often happens with e-commerce, education, SaaS, or lead gen websites using GA4, Google Ads, and conversion tags. If the cookie banner is implemented correctly, the consent/refusal state will determine which tags load and what data is sent.

When users refuse, is the data still usable?

Yes, but you need to be clear about what “usable” means. If users refuse certain processing purposes, you can still receive some minimal technical signals for aggregate measurement, while detailed data for personalized advertising is typically restricted.

Real-world example:

  • Users arrive on a landing page, refuse ads but accept measurement.
  • The system can record page_view and some conversions at an aggregated level.
  • Google Ads/GA4 can use modeling to estimate conversions that are not directly observed.
  • If users refuse measurement entirely, data for analytics will drop even more.

The important point: do not treat modeling as “measuring exactly like before.” Modeling fills gaps, but it always has latency, error margins, and depends on the quality of your baseline data.

User statusData collectedImpact on measurement
Full consentMore, fuller signalsReports closest to actual performance
Reject ads, accept measurementLimited data, may include aggregated signalsPartial measurement, with modeling
Reject allVery limited signalsSignificant reporting gaps, heavy reliance on inference

How does modeling in Consent Mode v2 work?

Modeling estimates missing data based on sessions from consenting users, aggregated signals, and the platform’s statistical models. It is especially useful when refusal rates are high, because without modeling, a substantial part of your conversion reporting would be “blind.”

But modeling is not magic. Three limits to remember:

  1. Not per-user raw data: you receive estimates at an aggregate level.
  2. Depends on sufficient data volume: if traffic is too low, models are less stable.
  3. Requires correct tag implementation: if consent signals are miswired, modeling can be skewed.

Example: a Google Ads campaign has 1,000 clicks, but 40% of users refuse. Reports may still attribute inferred conversions for the missing portion; however, CPA/ROAS may differ from the “raw” conversion counts in your backend CRM.

  1. Define your measurement goals:

    Be explicit whether you measure leads, purchases, signups, or calls. Each conversion type needs different consent configuration.

  2. Group tags by purpose:

    Separate tags for analytics, ads, remarketing, and third-party scripts.

  3. Implement the cookie banner correctly:

    Let users opt in/out per purpose—avoid vague bundling.

  4. Set consent signals before firing tags:

    Consent status must be read before sending events to platforms.

  5. Reconcile modeled data with backend data:

    Compare GA4/Google Ads against CRM, orders, or internal form submissions.

  6. Monitor refusal rates and data loss:

    If deny rates are high, improve banner content, UX, and transparency policies.

What data is still usable in Consent Mode v2?

Data that remains usable is usually the minimal data necessary for operations and aggregate measurement, as allowed by regulations and your consent configuration. However, “usable” still depends on your processing purposes, tag types, and system configuration.

Operationally, businesses should separate clearly:

  • Minimal technical data: page load, aggregated events, consent status.
  • Analytics data: used for performance reporting.
  • Advertising/personalization data: only activate when you have a legal basis and appropriate consent status.

Legally, collection and processing of personal data in Vietnam must comply with the Personal Data Protection Law (Law 91/2025/QH15), effective 01/01/2026. If your website stores cookies, uses trackers, or transfers data to advertising platforms, you need to design clear notice and consent mechanisms, store consent evidence, and prepare processes for data subject requests.

How do you measure well and still comply?

The right approach is not to “avoid consent,” but to design measurement with data minimization in mind. If your business runs an e-commerce or lead gen site, audit your cookie banner, CMP, consent logs, tag configuration, evidence retention periods, and DSAR processes.

If you use GA4, Google Ads, or third-party pixels, make sure to:

  • Provide clear notice of processing purposes.
  • Let users refuse each purpose group.
  • Store consent to demonstrate compliance.
  • Only send signals consistent with the consent status.

For SMEs, the hardest part isn’t drafting the policy—it’s wiring legal and technical correctly: banner, tag manager, event mapping, CRM, and reporting.

It reduces directly observed data when users refuse, but can partially compensate with aggregate-level modeling.
You may measure part of it depending on your consent setup and platform, but data will be more limited than with consent.
No. Modeling only estimates missing parts; it is not per-user raw records.
Provide clear notice, obtain appropriate consent, store evidence, and prepare processing under the Personal Data Protection Law (Law 91/2025/QH15); violations may be sanctioned under implementing decrees and, in serious cases, may face criminal enforcement.

If you need to implement a cookie banner, store consent evidence, or DSAR flows for your website/CRM, consent.vn can help you design a setup that both measures and is easy to verify for compliance.

Source: the Personal Data Protection Law (Law 91/2025/QH15) on thuvienphapluat.vn; Decree 13/2023/ND-CP on thuvienphapluat.vn; Ministry of Public Security (A05) on bocongan.gov.vn

Get started — set up in 5 minutes.

Need help with PDPL compliance?

Get started