GA4 Audit Checklist: 15 Things to Check Before You Trust Your Data

The short answer

Most GA4 setups I look at are quietly broken. Internal traffic isn't filtered. Key conversions aren't firing. Data retention is set to the default 2 months, so historical analysis is impossible. Cross-domain tracking is half-configured.

None of this announces itself. The reports load. The numbers move. They just aren't telling the truth. The point of this checklist is to catch the lies before you make decisions based on them.

Why GA4 audits get skipped

GA4 was forced on everyone in 2023 when Universal Analytics was sunset. Most teams migrated under deadline pressure, hit the basic setup, and moved on. The follow-up audit, the one that catches the broken pieces, never happened on most properties I see.

Then a few quarters pass. The team builds dashboards on the data. The CEO starts quoting it in board meetings. By the time someone notices that the lead-form conversion isn't actually firing, the numbers have been wrong for a year and the fix breaks the historical trend line everyone's been watching.

The 15-item audit

I run these in four groups. Configuration first, because if the property is misconfigured, none of the rest matters. Then conversion tracking, then data quality, then reporting connections.

Configuration (items 1 to 4)

1. 1Confirm the data stream is firing on every production domain. In Admin > Data Streams, verify the measurement ID is installed via Google Tag, GTM, or your CMS. Then load each domain in an incognito window with the GA Debugger extension to confirm hits actually arrive.
2. 2Review Enhanced Measurement settings. Page views, scrolls, outbound clicks, site search, video engagement, and file downloads are toggled on by default. Each one is useful. None are necessarily right. Confirm the toggles match what you actually want to measure.
3. 3Set Data Retention to 14 months. Default is 2 months on standard properties. That means anything older than 60 days is gone for explorations and reports that need user-level data. Admin > Data Settings > Data Retention. Change it. Today.
4. 4Decide on Google Signals deliberately. It enables demographics and interests but excludes some data from reports under thresholding rules. The right setting depends on your privacy stance and how much aggregate-only data you can tolerate. Pick on purpose, not by accident.

Conversion tracking (items 5 to 8)

1. 1Audit your Key Events list (formerly Conversions). Admin > Events. Confirm every event flagged as a Key Event is actually a real business outcome. Page views and scroll depth shouldn't be in this list.
2. 2Verify lead capture events fire. Submit each form on the site, place a test phone call if you track calls, start a chat session. Confirm the matching event appears in the DebugView in real time. Don't trust the historical reports until you've seen the events fire live.
3. 3Look for double-counting. A common pattern: a form_submit event AND a custom 'lead' event both fire on the same submission, both flagged as Key Events. Conversion totals double. Decide on one event per business outcome and turn off the duplicate.
4. 4Confirm conversion values are populated where applicable. If you assign a value to a lead or transaction, GA4 needs to receive it on the event. Without a value, all your revenue and ROAS calculations downstream are blank.

Data quality and filters (items 9 to 12)

1. 1Filter internal traffic. Admin > Data Streams > Configure Tag Settings > Define Internal Traffic. Add your office IPs and any team-member home IPs you care about. Then activate the Internal Traffic filter under Data Filters. Without this, your bounce rate, time on site, and conversion rate are warped by your own team.
2. 2Audit cross-domain measurement. If you have multiple domains in a single property (main site, shop subdomain, support portal, etc.), confirm cross-domain linking is configured under Data Streams. Otherwise the same user crossing domains gets counted as two users with two sessions.
3. 3Set referral exclusions for payment gateways. Stripe, PayPal, and any redirect-based checkout flow will appear as referral traffic, breaking your attribution. Add them under Data Streams > Configure Tag Settings > List Unwanted Referrals.
4. 4Spot-check the Acquisition reports for spam referrals. Self-referrals, bot traffic, and known-spam domains can all leak in. If something looks weird, look at the source/medium combo and the engagement metrics. Real traffic engages differently from bot traffic.

Reporting and integrations (items 13 to 15)

1. 1Link GA4 to Google Ads and Search Console. Both linkages live in Admin > Product Links. Without them, you can't pull paid-search performance into GA4 reports or see organic search queries inside GA4. Both are free and take five minutes.
2. 2Enable BigQuery export. Admin > Product Links > BigQuery Links. The export is free for standard properties (within reasonable thresholds) and unlocks user-level data, custom analysis, and long-term storage. It's the only way to do serious reporting outside of the GA4 UI.
3. 3Audit your Looker Studio or other reporting dashboards. Confirm every dimension and metric still exists in GA4 (some have been renamed or removed). Confirm date ranges aren't hardcoded. Confirm the data source connector is up to date. Stale dashboards are a faster way to make a bad decision than no dashboard at all.

What I find most often

Across the GA4 audits I've run, the same problems show up over and over.

1. 1Data retention still set to the default 2 months. Quietly the most common issue I find.
2. 2Internal traffic not filtered. The team's clicks inflate engagement, suppress bounce rate, and contaminate conversion tracking.
3. 3Lead-form conversion configured to fire on the form page load instead of the thank-you page. Looks fine until you realize the conversion count exactly matches the form-page traffic.
4. 4Stripe or PayPal redirects polluting the referral source for every checkout, breaking attribution.
5. 5Looker Studio dashboards built on GA4 dimensions that no longer exist or have been renamed, returning blanks or zeros that look like a traffic drop.

The scariest analytics problem isn't a missing report. It's a report that loads, looks normal, and is silently wrong. Audit before you trust.

The tools I actually use

Most of a GA4 audit happens inside GA4 itself. A few external tools speed things up. Once the property is clean, the GA4 reports that actually matter post covers which views to read every Monday.

• GA4 DebugView. Real-time event stream from your own browser. The fastest way to confirm events are actually firing the way you think they are.
• Google Tag Assistant (the new one, in tagassistant.google.com). Lets you walk through pages and see every tag firing, in order, with parameters.
• GA Debugger Chrome extension. Toggles debug mode for a session. Pairs with DebugView.
• Google Tag Manager preview mode. If you're using GTM (you should be), preview mode is the source of truth for what's firing and when.
• Looker Studio. Free dashboarding tool. Use it to surface the data so you actually look at it. Audits are easier when the metrics are visible.

Where this connects to technical SEO

A clean GA4 setup is half the foundation. The other half is making sure the site itself is healthy. If your analytics are pristine but the site is accidentally noindexed, you have great data on a problem nobody is finding.

Pair this audit with the technical SEO audit checklist. Run them together, in either order. They cover the two halves of the marketing-foundation question.

When to do this yourself, when to hire

You can run this checklist yourself if you have admin access to GA4, you're comfortable in GTM (or your team is), and you can dedicate a focused half day. Most properties take three to five hours to audit and another two to four to fix the issues found.

Hire it out if you've never set up GA4 from scratch, if you inherited the property from someone who's gone, or if you suspect the data has been wrong for a while and you need someone to write the migration plan to a clean baseline. The cost of bad analytics decisions usually outpaces the cost of an audit by an order of magnitude.

Getting started

If you want to work through this yourself, here's the order I'd follow.

1. 1Open GA4 Admin. Skim Data Streams, Data Settings, and Data Filters before doing anything else. Note what you see.
2. 2Change Data Retention to 14 months. Save.
3. 3Activate Internal Traffic filter (after defining your IPs).
4. 4Open DebugView. Submit every form, place every test action, confirm events fire as expected.
5. 5Open the Key Events list. Remove anything that isn't a real business outcome. Add anything missing.
6. 6Link Google Ads, Search Console, and BigQuery in Product Links.
7. 7Audit any dashboards you have. Replace renamed dimensions, fix date ranges, refresh data sources.
8. 8If any of this feels over your head, reach out. The full audit is part of every analytics engagement I run.

Most properties I audit (outdoor and DTC brands especially, where decisions get made off retailer-vs-direct revenue splits) have at least four issues from this list, and at least one is actively distorting decisions. The work isn't glamorous. It's just necessary. Get the data right, then start trusting it. Once the events are clean, the next step is usually a CRO audit to turn the working measurement into actual conversion lift.

Frequently asked questions

How long does a GA4 audit take?

Three to five hours of focused work for the audit itself, plus another two to four to fix what you find. Larger properties with multiple domains or e-commerce setups take longer. Most of the time goes into testing events in DebugView, not into reading the configuration screens.

What's the difference between a GA4 audit and a marketing-data audit?

A GA4 audit covers what's happening inside the GA4 property: configuration, events, filters, integrations. A marketing-data audit is broader. It includes GA4 plus your CRM, ad-platform tracking, attribution model, and the dashboards downstream. Start with the GA4 audit. The other layers depend on it being clean.

How often should I audit GA4?

Once on initial setup, after any major site change (redesign, migration, new conversion path), and at minimum once a year. Set a recurring calendar reminder. GA4 itself ships changes regularly (renamed dimensions, new defaults, deprecated features), and properties drift if nobody is watching.

Do I need GA4 if I have another analytics tool?

Probably yes, even if you also use Plausible, Fathom, Heap, Mixpanel, or another tool. GA4 integrates with Google Ads and Search Console in ways third-party tools don't. For paid-search and organic-search reporting in particular, GA4 is hard to fully replace. Keep GA4 clean and audited even if it isn't your primary analytics surface.

Keep reading