GA4 is powerful, but it’s also easy to misread if tracking, attribution, or conversions aren’t set up cleanly. Before I use GA4 to make budget decisions, report performance, or declare a campaign “working,” I run a quick audit.
Here are the 10 checks I verify first—plus the fast fixes that usually solve 80% of issues.
1) The Right Property + Data Stream (and it’s actually receiving data)
Sounds obvious, but it’s a common mistake: teams look at the wrong property, wrong stream, or a stream that stopped collecting.
- Confirm the correct GA4 property and web data stream
- Check Realtime while you visit the site and trigger a few events
2) Consent Mode / Cookie Banner impact (are you losing more data than you think?)
If consent is required, you may see reduced measurement (especially in EU/UK). That’s not “bad GA4”—it’s expected.
- Confirm your cookie banner + consent settings
- Note any big gaps between ad platform clicks and GA4 sessions
3) Cross-domain tracking (if you use a booking tool, Shopify checkout, 3rd-party forms, etc.)
Nothing breaks attribution faster than users jumping domains and getting counted as a new session (or “referral”).
- Identify all domains in the user journey (site → checkout → scheduler → portal)
- Look for self-referrals or unexpected referrers
4) Internal traffic filtering (and whether it’s accidentally filtering real users)
You want to exclude your team, agencies, and vendors. But I always confirm filters aren’t too aggressive.
- Check internal traffic rules and IP filters
- Verify you still see normal volume patterns after filters
5) Referral exclusions + payment processors (Stripe/PayPal) to prevent “fake” referrals
Payment gateways can “steal” credit from your campaigns if you don’t exclude them properly.
- Look for Stripe/PayPal/checkout domains showing up as referrers
- Confirm exclusions and cross-domain setup
6) Key events (conversions): defined correctly, deduplicated, and meaningful
GA4 conversions are only as good as what you mark as key events.
- Confirm the event fires only when a real conversion happens
- Watch out for “conversion inflation” (fires on page load, fires multiple times)
- Confirm naming consistency (especially if GTM has multiple versions)
7) Event parameters: capturing what you need to segment and optimize
It’s not enough to track “form_submit.” You need context.
- Form name / form ID
- Lead type or service category (if applicable)
- Content/page context
Quick win: add parameters so you can answer “which form?” and “which service?” without guessing.
8) UTM hygiene (or you’ll never trust channel performance)
Messy UTMs = messy channel reporting.
- Confirm consistent naming conventions (source/medium/campaign)
- Look for duplicates like facebook vs Facebook, or cpc vs paid
- Check for missing UTMs on email links and paid social
9) Attribution sanity check: do the trends match reality?
I don’t expect GA4 to match Google Ads 1:1—but I do expect directionally consistent trends.
- Compare GA4 sessions vs ad platform clicks (trend, not exact count)
- Check if “Direct” is suspiciously high (often a tagging or cross-domain issue)
- Check if paid traffic is being bucketed into Organic or Referral
10) Landing page + engagement checks (are you measuring the right outcomes?)
Before judging performance, confirm GA4 is reflecting user behavior accurately.
- Top landing pages make sense
- Engagement rates aren’t abnormally high/low due to broken events
- Key pages (thank-you pages, booking confirmations) are tracked correctly
My “Trust the Data” Minimum Standard
If these 3 are solid, I’m comfortable using GA4 for decision-making:
- Conversions are real (key events fire once, only on success)
- Attribution is sane (no self-referrals, cross-domain is clean, UTMs are consistent)
- Traffic is clean (internal filters + bots aren’t distorting the picture)
Quick Wins Recap (Most Common Fixes)
- Fix cross-domain tracking + referral exclusions
- Clean up UTMs and enforce a naming convention
- Deduplicate conversions and tighten trigger rules
- Add event parameters for segmentation
- Verify internal traffic filters are correct
Closing Thought
GA4 is a decision tool. If you don’t trust the inputs, you can’t trust the outputs.
If you want, I can run this checklist on your GA4 setup and deliver a short audit summary: what’s solid, what’s unreliable, and the highest-impact fixes to prioritize.