Understand why GA4 problems must be fixed bottom-up, and how to sequence your audit for maximum impact.
A GA4 account has four distinct layers, each depending on the one below it. This chapter explains the model and shows you the priority order that makes the biggest difference first.
A GA4 account has four distinct layers, each depending on the one below it. Fixing a problem in layer 3 without checking layer 1 and 2 first is how you end up with clean-looking reports that describe the wrong thing.
The four-layer model
Each layer depends on the one below. Work bottom-up.
Image placeholder
Layer dependency illustration
A visual showing how a problem in layer 1 (for example, the tag not firing on checkout pages) cascades upward through layers 2, 3, and 4, making each layer's data unreliable. Annotated arrows connect the root cause at the bottom to the corrupted metrics visible at the top, showing the analyst that what looks like a reporting problem is actually a collection problem.
Most analysts start at layer 4. They open the Acquisition report, look at the channel split, notice something odd, and try to fix it. But a broken channel split is usually a symptom of problems in layers 1, 2, or 3. Fix the foundations first and the reports often correct themselves.
Not all problems have equal impact. This is the sequence that addresses the highest-impact issues first, before you spend time on refinements that won't matter if the data underneath is unreliable.
Check real-time reports while browsing the site. If a key page type (product pages, checkout, blog posts) isn't sending events, you have a gap in your data that invalidates everything downstream. This is the most important check and takes less than 10 minutes.
If your team's own visits are still in the data, every metric — engagement rate, conversion rate, channel mix — is distorted. On a small-traffic site, a five-person agency team can represent 10-15% of sessions. Filter internal traffic and activate the filter (not just create it in Testing state).
GA4 defaults to 2 months of user and event data retention for Exploration reports. Once data ages out, it's gone permanently. Changing this to 14 months takes 30 seconds and prevents a problem that can't be fixed retrospectively.
Check which events are marked as key events. There should be 3-5 that represent real business outcomes. Verify that conversion counting is set correctly: "once per session" for lead forms (to avoid duplicate counts), "once per event" for purchases. Miscounted conversions are one of the most common inherited problems.
Image placeholder
Conversion counting diagram
A simple before/after showing how counting method (once per session vs once per event) affects the reported number of lead form conversions. Before: counting set to once per event, a user submitting 3 forms in one session generates 3 conversions. After: counting set to once per session, same behaviour generates 1 conversion. Numbers shown alongside to illustrate the scale of the discrepancy this creates over a month of data.
Payment gateways (Stripe, PayPal, Afterpay) often send a referral back to your site after the transaction, creating a phantom "referral" session that breaks the original source attribution. Add these domains to the referral exclusion list so the original session attribution is preserved through to the purchase event.
Once the data quality is solid, address how traffic and pages are grouped in reports. An Unassigned channel group above 5% usually means UTM gaps in paid campaigns. Missing brand vs non-brand split in organic means SEO wins and brand-driven sessions are being counted together.
If events carry custom parameters (a logged-in user ID, a content type, a product category), those parameters need to be registered as custom dimensions in GA4 Admin before they appear as reportable dimensions. Unregistered parameters sit in the data but can't be queried or filtered in standard reports.