Google Ads Conversions Vanished? Check This First

Google Consent Mode V2

I think of Google Consent Mode V2 as the measurement layer between a visitor’s privacy choice and Google’s systems. For a B2B service company, that layer affects reported conversions, Smart Bidding inputs, attribution, and the revenue credited back to campaigns. It is often one reason the number in Google Ads drifts away from what sales sees in the CRM, alongside the issues covered in Attribution windows explained: why your numbers do not match.

At its core, Google’s About consent mode documentation describes a framework where a cookie banner or consent management platform sends consent signals to Google tags. Those tags then change behavior based on what the visitor allowed. This matters most for teams using Google Ads, GA4, Google Tag Manager, and related Google measurement products, especially with visitors in the EEA or UK. Even outside those markets, I still care because privacy rules, browser restrictions, and measurement limits are moving in the same direction.

• ad_storage controls whether ad-related storage, such as cookies used for ads measurement, can be used.
• analytics_storage controls analytics storage for GA4 measurement.
• ad_user_data controls whether user data can be sent to Google for ads use cases.
• ad_personalization controls whether data can be used for personalized ads and remarketing.

In a B2B funnel, that affects more than a click and a form fill. An ad click leads to a landing page, the page sends a default consent state, the visitor makes a choice, tags adjust, a form or booked call may happen, the CRM records the lead, sales qualifies it, and closed revenue may later be imported back into Google Ads. When consent mode is clean, I can still read that funnel. When it is not, the middle of the journey becomes hard to trust. If you want a narrower lead-gen view, Consent Mode v2 for Lead Gen: Getting Measurable Signals Without Guesswork is a useful companion.

How consent mode works

Once I strip away the jargon, the logic is simple. The site sets a default consent state when the page loads. In regions where consent is required, that usually means denied by default. The banner appears, the visitor makes a choice, and the consent state updates. Google tags read that state and adjust their behavior.

If consent is granted, tags can measure more fully. If consent is denied, consent-aware tags restrict what they store and send. In advanced setups, they can still send cookieless pings. Those pings are easy to overlook, but they matter because they provide limited, privacy-conscious signals that can support modeled conversions.

This is usually where Google Tag Manager and the consent platform work together. The banner captures the choice, and GTM passes the right consent state to GA4, Google Ads, and related tags. The principle is straightforward. The execution is where most problems start.

Basic consent mode

Basic consent mode is the stricter approach. Google tags wait for user consent before they fire. If the visitor declines, the tags stay blocked and Google receives little or nothing from that visit.

I understand why many teams choose this route. It feels conservative, it is easier to explain internally, and it is usually simpler to launch. The tradeoff is thinner measurement. In lead gen accounts, that often means fewer recorded Google Ads conversions, weaker GA4 coverage, smaller remarketing pools, less signal for Smart Bidding, and a wider gap between ad platform reporting and CRM reality.

Basic mode still has value. I would take it over having no consent mode at all, because at least the system receives a structured consent state. But when an account depends on steady learning from qualified leads, the measurement loss becomes hard to ignore.

Advanced consent mode

Advanced consent mode also starts from denied by default, but the tags load earlier and can send cookieless pings while consent remains denied. If the visitor later grants consent, measurement becomes fuller. If the visitor says no, Google still receives limited signal that can support modeling.

That is why advanced mode usually preserves more measurement than basic mode. It does not create perfect tracking, and I would not present it that way. What it does is reduce silence. In B2B accounts where I rely on qualified lead volume, that difference can materially improve reporting, attribution, and Smart Bidding inputs.

The catch is setup quality. The consent platform has to send the right state, Google Tag Manager has to fire defaults early enough, and GA4 and Google Ads tags have to respect those states. One weak link can undo the benefit. When legal review and regional rules allow it, I usually see advanced mode as the stronger measurement choice. Google’s guide to Set up Basic or Advanced Google Consent Mode is the best baseline if you need the official implementation logic.

Conversion tracking

Conversion tracking is where this stops being a technical discussion and starts affecting revenue decisions. Consent signals influence Google Ads conversions, GA4 events, remarketing lists, attribution paths, offline conversion imports, and the final revenue view in a CRM-connected funnel.

A plain B2B example is more useful than theory. A prospect clicks a Google ad for a specialized service, lands on the site, declines analytics and ad cookies, but still submits a demo request because the offer matches the need. Sales books the call, qualifies the lead, and closes a large deal several weeks later. The important question is not whether the campaign worked. It did. The real question is how much of that path remains visible.

With no consent mode, Google Ads may miss the original conversion or record only fragments of it. GA4 can lose session continuity. Remarketing lists may never include the visitor. If the CRM import depends on identifiers that were never captured, closed revenue may not make it back to the ad platform either.

With advanced Consent Mode V2 and a sound implementation, the path is still imperfect, but it is much more readable. Google can combine consent signals with cookieless pings for modeled conversion reporting, and if lead capture plus CRM import are wired correctly, more of the journey becomes measurable. That is why I push teams to define stages clearly with How to define meaningful conversion events for B2B and to connect sales outcomes through Offline Conversion Imports: The Only Signals Google Ads Should Optimize For.

This is also where the four signals stop feeling abstract. ad_storage affects click storage and conversion linking. analytics_storage affects GA4 session and event continuity. ad_user_data shapes whether certain data can be sent to Google for ads measurement use cases. ad_personalization affects remarketing and audience use.

I understand why executives get impatient here. They are not asking for a lesson on privacy mechanics. They want the number in Google Ads to sit closer to the number in HubSpot or Salesforce. Consent mode will not remove every mismatch, but a clean setup can narrow the gap instead of letting it widen.

Consent mode modeling

Modeled conversions are Google’s estimated conversions based on observed behavior plus limited signals from users who did not grant full consent. In plain terms, Google tries to fill in some of the missing pieces when direct measurement is incomplete.

I do not expect modeled lift to appear immediately. In many accounts, it takes at least seven full days before any pattern becomes visible, and lower-volume B2B accounts often need longer. Modeling works best when traffic is steady, consent signals are accurate, tags fire in the right order, region logic is correct, and conversion actions are clearly defined. If that foundation is loose, start with a clean measurement plan such as B2B Conversion Tracking Checklist: GA4 Events That Matter.

Just as important, modeling has clear limits. It cannot repair a broken CRM import, recreate sales stages that were never tracked, or rescue a tag that never fired. It also cannot compensate for weak campaign structure. I think of it as gap reduction, not a substitute for implementation quality.

That distinction matters in lead generation. A B2B company may only generate a few dozen primary leads a month, yet a handful of missing conversions can still distort budget decisions. When volume is low and deal value is high, small reporting errors have oversized consequences.

No consent mode

No consent mode means Google receives no structured consent signals at all. When I see that setup, I usually see smaller audience pools, lower reported conversion counts, shakier Smart Bidding, and weaker visibility from click to revenue.

The important nuance is that missing reporting does not always mean missing revenue. Campaigns can still influence pipeline while the platform sees less proof of it. The problem is that the algorithm reacts to the signals it receives, not to what the sales team suspects happened. That is why doing nothing often hurts performance twice: first by hiding conversions, and then by feeding weaker data back into bidding.

Missing consent mode

Sometimes consent mode exists on paper but fails in practice. I see that almost as often as I see it missing entirely. A late default signal, conflicting banners, incorrect regional rules, or tags firing in the wrong order can distort reporting even when everyone assumes the setup is fine.

The symptoms are usually familiar. Conversion counts drop right after a banner change. GA4 and Google Ads disagree more than usual. Remarketing audiences shrink. Consent rates look stable, but ad platform reporting falls sharply. When I see that pattern, I assume setup trouble before I assume market trouble.

To verify what is happening, I use Tag Assistant to inspect consent default and consent update events, then I watch browser network requests before and after banner interaction. I also test accepted and rejected states, check how region logic behaves, and confirm that GA4, Google Ads, Floodlight if it is in use, and Conversion Linker all react to the same consent state. If the journey crosses domains or subdomains, I verify that consent remains consistent there too. Google’s checklist to Verify consent mode implementation (web) is a good starting point, and I usually pair it with the same discipline I use in Conversion sanity checks before you scale ad spend.

Consent mode setup

A good setup is not magic. I usually reduce it to sequence, testing, and discipline.

I start with a consent management platform that can send Google-compatible consent signals and handle the regions involved. Then I map the banner choices to the four consent signals so each user action leads to a clear technical state. In Google Tag Manager, I make sure consent defaults fire before other Google tags, and I configure consent updates so the choice changes the state immediately. After that, I confirm that GA4, Google Ads, Conversion Linker, and any Floodlight tags all read the same signals.

This is a common place to use GTM with a CMP, and that pairing works well when the handoff is clean. Built-in Google tags generally support consent behavior, while custom tags often need extra checks. I also make sure form tracking, call tracking, and the CRM handoff still work in both consented and denied scenarios. Regional behavior needs testing, especially when rules differ by country, and I keep watching reporting for a few weeks after launch before I judge the impact.

This is also where I often find a deeper measurement problem. Many firms track only form fills, not the stages that follow. If the sales cycle runs through booked calls, qualified meetings, proposals, and closed-won deals, the measurement plan should reflect that path. Consent mode improves signal quality, but it cannot add business stages that were never captured.

Revenue recovery

I use the phrase “revenue recovery” carefully here. Often the first thing that improves is reported revenue, not actual revenue. That still matters. Better reporting gives Smart Bidding better inputs, and better inputs can lead to better campaign decisions over time. So the first gain is usually visibility, and the second gain may become real commercial performance.

If I want the plainest version of the argument, it is this: many B2B firms do not have a traffic problem so much as a measurement problem. When Google sees too little, it bids too cautiously or learns from the wrong patterns. When it sees more of the funnel, it usually makes better decisions.

There is one important detail here. Better lead tracking can be paired with offline conversion imports and, where relevant, Enhanced Conversions for leads. That can improve the connection between ad clicks and CRM outcomes, but it does not replace consent mode. It works better alongside consent mode, not instead of it.