70% of Google Ads accounts have broken tracking. That's not speculation - it's what we found auditing 20+ brands between 2023 and 2025. The average brand wastes 23% of ad spend because of data problems.
Here's what makes tracking failures so dangerous: most brands don't know they're broken. GA4 says one thing. Google Ads says another. Shopify says something else. Every number looks plausible. None of them agree.
This diagnostic covers Layer 1 - the foundation of your tracking stack. If anything here fails, every layer above it is unreliable. Your attribution model doesn't matter if your accounts aren't linked. Your Enhanced Conversions don't matter if consent mode is misconfigured.
Start here. Fix here. Then audit the rest.
Think of your tracking as a 5-layer stack. Each layer depends on the one below it.
Layer 5: Optimization signals (bidding data, audience signals) Layer 4: Attribution (cross-platform alignment, windows) Layer 3: Conversion actions (what you're tracking, values) Layer 2: Tag implementation (GTM, pixels, event firing) Layer 1: Foundation (account linking, consent, base setup)
If Layer 1 is broken, Layers 2-5 are corrupted by default. You can have the best attribution model and the most sophisticated audience signals - none of it works if the foundational connections aren't right.
That's why we always audit bottom-up. And that's why this diagnostic starts here.
Open GA4 > Admin > Product Links > Google Ads Links.
Pass: Your Google Ads account ID appears and shows "Linked." Fail: No link, or wrong account linked.
Impact if broken: No audience sharing between platforms. No cross-platform reporting. Remarketing audiences from GA4 won't be available in Google Ads. You're running blind on how paid traffic behaves after it reaches your site.
While in GA4 > Admin, verify the property ID matches the one on your website.
Pass: The linked GA4 property is the correct, active property for your site. Fail: You're linked to an old Universal Analytics property, a test property, or the wrong domain's property.
Impact if broken: All data flows to the wrong destination. Conversion data, audience data, and reports are either missing or from the wrong source. We've seen brands run for months linked to a staging property while their production data went nowhere.
Open Google Ads > Tools & Settings > Linked Accounts > Google Merchant Center.
Pass: Your Merchant Center account appears and shows an active link with a healthy product feed. Fail: Not linked, or linked but showing feed errors or disapproved products.
Impact if broken: No Shopping campaigns. No product-based PMax campaigns. If linked but the feed has issues, products are silently excluded from auctions. You could have 1,000 products and only 400 eligible to show.
Note: Skip this check if you're not running e-commerce or Shopping campaigns.
Open Google Tag Manager > Admin > User Management.
Pass: Your Google Ads account has admin or publish access to the GTM container. Fail: No access, or read-only access.
Impact if broken: You can't modify tracking tags, update conversion configurations, or deploy fixes without going through someone else. For agencies and contractors, this is a common bottleneck that delays critical tracking fixes by days or weeks.
Open Google Ads > Admin > Account Settings.
Pass: Auto-tagging is turned on. Fail: Auto-tagging is disabled.
Impact if broken: Auto-tagging appends a GCLID parameter to your landing page URLs. This is the primary mechanism for Google Ads to track which clicks convert. Without it, conversion attribution breaks down. Google can't tell which keyword, ad, or campaign drove the sale.
Open your GTM container. Look for consent initialization tags.
Pass: Consent Mode v2 is implemented. Consent banner is connected to Google tags. The ad_storage and analytics_storage parameters are set correctly based on user consent choices.
Fail: No consent mechanism in place, or basic consent banner not connected to Google tags.
Impact if broken: Required for EU/EEA/UK since March 2024. Increasingly important for US state privacy laws. Without Consent Mode v2, every user who declines cookies becomes invisible to your tracking. Google uses consent signals to model conversions from opt-out users - without it, your conversion data has a gap that grows as privacy awareness increases.
Quick test: Open GTM Preview mode. Load your site. Check that consent tags fire before any Google Ads or GA4 tags in the firing sequence.
With GTM Preview still open, interact with your consent banner (accept and decline cookies).
Pass: Accepting cookies triggers Google tags to fire normally. Declining cookies shows tags in "blocked" or "consent denied" state but still sends consent pings. Fail: Tags fire regardless of consent choice, or no tags fire at all after declining.
Impact if broken: If tags fire regardless of consent, you're violating privacy regulations. If nothing fires after declining, you're losing all conversion data from privacy-conscious users. Consent Mode v2's modeling feature only works when consent signals are properly passed.
Still in GTM Preview mode. Navigate through your purchase funnel while monitoring tag firing.
Pass: Consent-dependent tags only fire after consent is granted. Non-consent tags (basic analytics pings) fire with appropriate consent parameters. Fail: Tags fire out of order, or consent tags don't influence other tags' firing behavior.
Impact if broken: Firing order matters. If your Google Ads conversion tag fires before the consent initialization tag, it runs without consent context. This creates inconsistent data - some sessions tracked with consent, others without - making your conversion data unreliable.
Click on one of your Google Ads (or simulate by adding ?gclid=test123 to your landing page URL).
Pass: The GCLID parameter remains in the URL through your entire site, including any redirects. Fail: The GCLID disappears after a redirect, or your site strips URL parameters.
Impact if broken: Near-total attribution loss. Google literally can't connect any click to any conversion. Found in 19% of accounts we audit. Common culprits: Cloudflare Page Rules that strip URL parameters, server-side redirects that drop query strings, URL canonicalization plugins, and CDN configurations.
How to test: Add ?gclid=test123 to your homepage URL. Navigate to a product page. Navigate to checkout. If at any point the URL no longer contains the gclid parameter, something in your redirect chain or site configuration is stripping it.
If your checkout lives on a different domain (like Shopify's checkout.shopify.com), this check is critical.
Open GA4 > Admin > Data Streams > Select your stream > Configure Tag Settings > Configure Your Domains.
Pass: All domains in your purchase flow are listed. Test purchase shows the session persisting across domains in GA4 Realtime. Fail: Checkout domain not listed. Session resets at domain boundary.
Impact if broken: Found in 28% of accounts. When the session breaks at checkout, every conversion looks like "Direct" traffic in GA4. Google Ads can't attribute the sale to the click. 20-40% of conversions become invisible to Smart Bidding.
Shopify-specific: If you're on Shopify, add checkout.shopify.com to your cross-domain configuration. Then complete a test purchase and watch GA4 Realtime to verify the session doesn't reset at checkout.
Skip this check if your entire purchase flow stays on a single domain.
Count your passes out of 10 (skip any that don't apply to your setup).
| Score | Status | Action |
|---|---|---|
| 9-10 | Solid Foundation | Your tracking base is reliable. Move to Layer 2 (tag implementation). |
| 7-8 | Minor Gaps | Fix the failed checks before optimizing anything else. Each gap corrupts data above it. |
| 5-6 | Foundation Issues | Your Smart Bidding is operating on partial data. Fix these before spending more. |
| Below 5 | Critical | Your conversion data is unreliable. Budget decisions based on these numbers are guesses. |
If you have multiple failures, fix them in this order:
This diagnostic covers Layer 1 - the 10 foundation items. The Google Ads AI Agentic System ($4,997+) covers all 5 layers with 50 items total:
It also includes platform-specific troubleshooting for Shopify, WooCommerce, and BigCommerce, plus the 5 most common tracking failures ranked by frequency with step-by-step diagnostics.
Your foundation determines everything above it. If Layer 1 passes, the rest of the audit is about precision. If Layer 1 fails, the rest of the audit is about survival.