GA4 Setup Checklist: 12 Things Most Businesses Get Wrong

I've audited a lot of GA4 properties, and the same handful of issues show up again and again — not exotic edge cases, just basic setup steps that got skipped or half-finished. Here are the twelve I see most often.

Data quality basics

• Internal traffic isn't filtered. Office and agency visits are still inflating every report.
• No cross-domain measurement when checkout or booking lives on a different domain, splitting sessions in two.
• Referral exclusions aren't set for payment providers, so Stripe or PayPal shows up as a traffic source instead of the original channel.
• Bot traffic filtering left on defaults without checking whether it's actually catching what hits your site.

Tracking configuration

Enhanced measurement settings match what you actually want captured

GA4 ships with a set of enhanced measurement toggles that automatically capture certain interactions in a fixed format. Left on defaults, they're a reasonable starting point — but adjusting them to match what you actually want tracked is worth the ten minutes it takes.

First step: in GA4 Admin → Data Streams → Enhanced measurement, check which toggles are actually on against what you assumed was being tracked.

Consent Mode configured correctly — and not silently dropping data

Consent Mode is a more recent, privacy-focused part of the tracking stack, and it's a common source of incomplete or missing data when it isn't wired up correctly. It needs a careful eye to confirm it's actually working as intended, not just installed.

First step: go to Consent Settings in GA4 and check whether the status shows as active. If it doesn't, you'll need some changes to your tracking setup.

No duplicate tracking (old snippets, double Tag Manager containers)

There are several ways to add tracking to a site, and Google has built simpler and more advanced methods depending on your needs — great for reducing implementation overhead, but it also means different people can end up adding different tracking methods to the same site over time. The result is duplication.

First step: load a test URL on your site once and see how many sessions appear against it. Use a URL that's no-indexed and blocked from search engine and AI/bot crawler traffic, so you're only counting your own visit.

E-commerce events fire with correct parameters (value, currency, items)

E-commerce tracking data is essential for properly analysing a transactional site, but missing value or item data is a common issue — and it makes reconciling GA4 against your actual sales figures very difficult.

First step: compare your GA4 e-commerce data against your inventory data and check whether all the data points you'd expect are actually present.

Data governance & validation

UTM parameters used consistently across campaigns

UTM parameters are genuinely useful for understanding which campaigns, creative and traffic sources are working — but they're often applied inconsistently across a team, which leaves you with incoherent groupings that make analysis unreliable. The right fix depends on your team structure and how you prefer to tag campaigns.

First step: pull last month's channel report and count how many distinct source/medium values exist for what should be three or four channels. More than ten or so means you've got a tagging-discipline problem, not a technical one.

Bot/spam traffic filtered out

Bot and spam traffic — often AI crawlers or scrapers gathering content or doing competitor research — is a common and hard-to-spot issue in analytics data. It's not always easy to exclude, but doing so meaningfully cuts down the overhead of cleaning up data that's been skewed by unexplained traffic spikes.

First step: review your traffic for unusual spikes, then drill into any specific dates that look off.

GA4 numbers reconcile against a second source (Search Console, ad platform, CRM)

Comparing GA4 against Search Console, ad platforms or your CRM is a useful sanity check, but it's rarely a clean exercise — each of those sources has its own definitions and exclusions that don't map neatly onto web analytics. Don't expect the numbers to match exactly; expect them to be in the same ballpark.

First step: compare last week's GA4 sessions against Search Console clicks for the same date range. A wide gap is worth investigating; a small one isn't.

A documented owner and change log exist for the tracking setup

Untangling an existing tracking setup can take hours of testing and configuration review — a problem a simple change log largely prevents, as long as it's actually kept up to date.

First step: if you're using Google Tag Manager, check the version history — it gives you a ready-made timeline of what's changed and when.

None of these are hard to fix individually. The problem is they compound — by the time you've got four or five of them at once, the whole property stops being trustworthy.

If you're not sure which of these apply to your setup, the fastest way to find out is to open GA4's DebugView, browse your own site, and watch what actually fires. Most of the list above becomes obvious within about ten minutes of doing that.