Your Google Ads conversion numbers went sideways sometime in the last year and you can't tell if it's your campaigns, your website, or something in the tracking pipe itself. The real answer, for most small businesses running ads in 2025, is a plumbing problem: Consent Mode v2 rewired how Google is allowed to measure conversions, and most sites never finished the install.
This guide walks through what changed, why your numbers look off, and exactly how to rebuild tracking so Google Ads reports what's actually happening on your site.
In This Guide
- Consent Mode v2 Changed What Google Ads Can Measure
- Why Your Conversion Numbers Look Wrong Without the New Signals
- Step 1: Audit What's Actually Firing on Your Site Today
- Step 2: Implement the Two Consent Mode v2 Signals Correctly
- Signal Comparison: Consent Mode v1 vs Consent Mode v2
- Step 3: Connect Your Consent Management Platform to Tag Manager
- Step 4: Verify Tracking With Tag Assistant and Ads Diagnostics
- Step 5: Expect a Gap Between Observed and Modeled Conversions
- The Silent Failure Mode Nobody Checks: Tracking Beacons Eating Performance
- How Las Vegas Small Businesses Can Validate Their Own Setup
- FAQ
Consent Mode v2 Changed What Google Ads Can Measure
Google required Consent Mode v2 signals for any advertiser serving ads to users in the European Economic Area starting March 2024, according to support.google.com. If you only run campaigns in the US Southwest, that deadline may feel like someone else's problem, but the ripple effect hit accounts everywhere: Google Ads tightened how it reports conversions across the board, and plenty of sites without the updated signals have seen conversion counts drop or stop populating entirely in certain regions.
Consent Mode v2 added two new signals on top of the original pair:
- analytics_storage and ad_storage (the original Consent Mode v1 flags)
- ad_user_data, new in v2
- ad_personalization, new in v2
If your tag setup only knows about the first two, Google Ads is missing half the picture it needs to measure anything reliably.
Why Your Conversion Numbers Look Wrong Without the New Signals
Here's the mechanism, in plain terms. When a visitor declines cookies, Google Ads can no longer rely on a stored click ID to tie a conversion back to the ad that drove it. That's the whole point of consent: no cookie, no persistent identifier.
Google fills that gap with modeled conversions, statistical estimates built from aggregated, consented data rather than a direct match. Think of it as Google saying "based on everyone who did consent, we estimate X percent of the non-consenting group also converted."
The catch: if your tags still send only the old v1 consent parameters, Google Ads has no signal to model from, and conversions simply under-report. It isn't that your ads stopped working; it's that the system reporting on them went blind in one eye.
This is the single most common reason small business owners tell us "our conversions tanked and we didn't change anything." You didn't. Google did.
Step 1: Audit What's Actually Firing on Your Site Today
Before fixing anything, find out what's actually installed. Open Google Tag Manager's preview mode and confirm whether gtag.js, the Google Ads tag, and any server-side container are even present on the page.
While you're in there, build a quick inventory:
- List every third-party pixel firing on key pages: Meta, LinkedIn, TikTok, and any analytics snippets stacked on top of each other.
- Note whether consent signals are being read from a real consent management platform (CMP) or hardcoded as "granted" by default, a shortcut that's common and genuinely risky from a compliance standpoint.
- Flag any tags firing before a consent decision has even been made.
If you've never run conversion tracking through GA4 before, our guide to setting up conversion tracking in GA4 is a good companion read for this audit step, since GA4 and Google Ads conversions often share the same underlying tag infrastructure.
Step 2: Implement the Two Consent Mode v2 Signals Correctly
Once you know what's missing, add it. The developers.google.com documentation on the gtag consent API defines the two new parameters precisely:
- ad_user_data controls whether user data can be sent to Google for advertising purposes at all.
- ad_personalization controls whether that data can be used to build personalized or remarketing audiences.
Set sensible defaults, denied until consent is actually given, and update them dynamically the moment a user interacts with your cookie banner. Don't set these to "granted" by default just to make your dashboards look fuller; that defeats the purpose of consent mode and can misrepresent what you're actually allowed to do with the data.
Signal Comparison: Consent Mode v1 vs Consent Mode v2
| Signal | v1 | v2 |
|---|---|---|
| ad_storage | Yes | Yes |
| analytics_storage | Yes | Yes |
| ad_user_data | Not present | Required |
| ad_personalization | Not present | Required |
| Modeled conversions | Limited | Expanded, built on the two new signals |
The table makes it obvious why incomplete installs under-report: two of the four required signals simply didn't exist before v2, and Google's modeling depends on having all four.
Step 3: Connect Your Consent Management Platform to Tag Manager
Most CMPs, OneTrust, Cookiebot, and other Google-certified alternatives, push consent state into GTM's dataLayer automatically once configured correctly. Use GTM's native Consent Mode templates instead of writing custom JavaScript, unless you have a specific reason not to, like a non-standard CMP or a server-side container with custom routing.
Don't just check that the banner displays and call it done. Test that the actual consent state changes in the dataLayer when a visitor clicks accept versus reject. A banner that looks right but doesn't fire the underlying signal is worse than no banner at all, because it gives a false sense of compliance.
Step 4: Verify Tracking With Tag Assistant and Ads Diagnostics
Verification has three parts:
- Run Google's Tag Assistant Chrome extension against your live site and confirm the Google Ads conversion tag fires with the correct consent parameters attached.
- Check the Diagnostics tab inside your Google Ads conversion action for warnings about missing consent signals. Google surfaces this directly; it's not hidden.
- Run a real test conversion and confirm it appears in Google Ads within 24 to 48 hours, understanding that modeled data can lag further behind.
If you're also running Performance Max campaigns, it's worth cross-checking this against our step-by-step Performance Max setup guide, since PMax leans heavily on conversion data quality to optimize bidding.
Step 5: Expect a Gap Between Observed and Modeled Conversions
Two terms you'll see constantly in Google Ads reporting from here on:
- Observed conversions: ones Google can directly attribute to a click.
- Modeled conversions: the statistical fill-in from aggregated, consented signal.
A sudden shift in the ratio between the two right after launch is normal. It doesn't automatically mean tracking is broken; it usually means the new signals are working and Google is routing more conversions through the modeling path than before.
Give the system at least one to two full weeks of consistent traffic before judging whether modeling has stabilized. Pulling the plug after three days is the most common mistake we see here.
For more context on what Consent Mode v2 changes structurally for advertisers, our earlier piece on Consent Mode v2 for small business advertisers covers the compliance side in more depth.
The Silent Failure Mode Nobody Checks: Tracking Beacons Eating Performance
Here's the part most consent mode guides skip entirely: fixing your tracking can quietly break your site speed.
Adding conversion pixels, remarketing tags, and a CMP script all at once is exactly the kind of change that degrades Core Web Vitals without a single code deploy touching your actual site. In one of our client audits, a mobile Lighthouse score fell from 96 to 41 over six weeks with almost no deploys to the codebase itself; the cause traced back to a stack of five third-party tracking beacons layered on top of each other, not a regression in the underlying code.
That's an easy trap to fall into right when you're doing exactly the right thing by fixing your consent signals. The fix is simple but easy to skip:
- Audit your tag stack for redundant pixels every time you add a new conversion tool.
- Measure impact with field data over time, not a single local Lighthouse run that only captures one moment.
- Remove duplicate analytics snippets that accumulated from old CMPs or abandoned tools.
If your Core Web Vitals haven't been checked recently, run through our Core Web Vitals checklist alongside this tracking cleanup. They're the same audit, really, just pointed at two different symptoms of the same tag bloat.
How Las Vegas Small Businesses Can Validate Their Own Setup
Once everything's implemented, prove it to yourself with a real-world test:
- Run a real test purchase or lead form submission on your own site from a Las Vegas IP and confirm it shows up in Google Ads as an observed conversion.
- If you run both Google Ads and Meta campaigns, check that your CMP is passing consistent consent state to both platforms, not just Google. A consent gap on one platform but not the other is a common half-finished install.
- Pair conversion tracking fixes with a broader technical audit. A slow or poorly tagged site undermines ad spend no matter how clean the tracking math is underneath it.
If page speed issues from the tag stack turn out to be render-blocking rather than just beacon bloat, our guide on diagnosing render-blocking scripts walks through isolating the specific script causing the slowdown.
Need help auditing your current tag stack rather than guessing at what's firing? Our paid ads management services include a tracking audit as a standard first step, precisely because we keep finding broken consent signals hiding behind otherwise solid campaigns.
FAQ
Do I need Consent Mode v2 if my business doesn't serve ads in Europe? Google requires Consent Mode v2 signals specifically for advertisers reaching users in the European Economic Area, per support.google.com. If your campaigns only target the US Southwest, you aren't required to implement it today, but Google has signaled broader rollout plans, so setting it up now avoids a scramble later.
What's the difference between ad_user_data and ad_personalization signals? ad_user_data controls whether user data can be sent to Google for ad measurement at all. ad_personalization controls whether that data can be used to build remarketing audiences or personalized ads. A visitor can consent to one and deny the other, so both signals need independent handling in your tag setup.
Will my Google Ads conversion numbers drop after implementing Consent Mode v2? Observed conversions may appear to drop because some are now reported as modeled conversions instead of direct matches. Total reported conversions, observed plus modeled, typically stabilize within one to two weeks once Google has enough consented data to build an accurate model.
Can I use Google Tag Manager's built-in consent templates instead of custom code? Yes. GTM ships native Consent Mode templates that most consent management platforms integrate with directly, which is simpler and less error-prone than hand-coding the gtag consent API. Custom code is worth it only if you have a non-standard CMP or a server-side container with specific routing needs.
How long does Google take to model conversions after consent signals go live? Google needs a meaningful volume of consented, directly-observed conversions before its model can reliably estimate the rest. Most accounts see modeling stabilize within one to two weeks of consistent traffic; low-volume accounts may take longer to reach a reliable baseline.
Your next move: pull up Tag Assistant on your own live site today and check whether ad_user_data and ad_personalization actually appear in the console output when a test visitor accepts cookies. If they don't, you've found the leak.