How to Know If Your GA4 Is Tracking Correctly

Most Google Analytics 4 properties have at least one significant tracking problem. Not a cosmetic one. A problem that changes the numbers your team reports, the campaigns your budget flows toward, and the conclusions your leadership draws from a quarterly deck.

The uncomfortable part is that GA4 will never tell you. It does not flag a conversion firing twice. It does not warn you that a tag is loading on every page instead of the thank you page. It reports whatever it receives, confidently, and your team plans around it.

If you have ever looked at a report and thought the number seemed off, this is how you find out whether your instinct was right.

The most common signs your GA4 is misconfigured

Broken tracking has a recognizable shape. These are the symptoms that show up in reports before anyone opens a container.

Your conversion count does not match your back end. GA4 says 340 purchases last month. Shopify says 210. Some variance between platforms is normal because of attribution windows and time zones. A 60 percent gap is not variance, it is duplication.

Conversions arrive in suspiciously round clusters. If your form submissions climb at exactly the same rate as your sessions, the event is probably firing on page load rather than on an actual submission.

Your direct traffic is enormous. Direct should be people typing your URL or using a bookmark. When it becomes your largest channel, source data is usually being lost somewhere between the ad click and the conversion.

Engagement metrics look impossible. Average engagement time of four seconds, or a bounce pattern that contradicts what your heatmaps show, usually means duplicate page_view events are inflating the denominator.

Google Ads and GA4 disagree loudly. Two platforms counting the same action should land within a reasonable range of each other. When Ads reports triple what GA4 does, one of them is counting something the other is not.

Nobody can explain an event name. If your events list contains three variations of the same action, added by three different people over two years, your data has been fragmenting quietly the entire time.

How to check your conversion events in GA4

Start with the events themselves, because everything downstream depends on them.

Open Admin, then Events, and read the full list. Ask a blunt question of each one: do I know what user behavior triggers this, and do I know who added it? Any event that fails both halves of that question is a candidate for removal.

Next, open Admin, then Conversions, and check which events are marked as key events. This list should be short. If a dozen events are marked as conversions, your reporting cannot distinguish between someone buying and someone scrolling.

Then use Realtime. Open your site in another tab, complete the action yourself, and watch whether the event appears once or twice. Do this for every conversion you actually care about. It takes fifteen minutes and it catches the majority of duplication problems.

Finally, compare a full month of GA4 conversions against the source of truth. Your CRM, your ecommerce backend, your booking system. Write both numbers down. That gap is the size of your problem.

How to use GTM preview mode to verify tag firing

Realtime tells you an event arrived. Preview mode tells you why.

In Google Tag Manager, click Preview and enter your site URL. Tag Assistant opens a debug session alongside your site. As you click through, the left panel logs every event and the right panel shows which tags fired, which did not, and why.

Three things are worth checking carefully.

Look at what fired on page load. Any conversion tag appearing here is misconfigured. Conversions should be tied to a specific interaction, not to a page rendering.

Check the Tags Not Fired list. This is the more useful panel. A tag that never fires is invisible in reports, so nobody notices it is missing. Reviewing this list surfaces tracking you thought you had.

Inspect the data layer on each event. Click an event and open the Data Layer tab. If your purchase event is passing an undefined value, or a transaction ID that repeats across sessions, your revenue reporting and your deduplication are both compromised.

Walk your key paths this way. Homepage to product to cart to checkout. Landing page to form to thank you page. Fifteen minutes per path will tell you more than a week of staring at reports.

What to look for in your data streams and property settings

Plenty of problems live above the event layer, in settings nobody revisits after setup.

Internal traffic filters. Admin, then Data Streams, then Configure tag settings. If your team, your agency, and your developers are not excluded, a meaningful share of your sessions is your own staff.

Unwanted referrals. Payment processors and booking tools that redirect users off site and back will break the session and steal attribution credit unless listed here. Stripe and PayPal are the usual offenders.

Session timeout. The default is 30 minutes. If someone lengthened it, your session counts and engagement rates are not comparable to any benchmark.

Data retention. The default is two months. Two months. Change it to fourteen, which is the maximum, and do it today, because retention is not applied retroactively.

Cross domain configuration. If your site spans multiple domains, and they are not configured here, every hop between them starts a new session with a new source.

When to fix it yourself and when to bring someone in

Some of this is genuinely a do it yourself job. Extending data retention, adding internal traffic filters, unmarking events that should never have been conversions, listing payment processors as unwanted referrals. These are settings changes with a clear right answer and low risk.

The line moves when the fix touches the container. Rebuilding conversion tracking, implementing a data layer with your developers, resolving duplication caused by both a hardcoded tag and a GTM tag firing at once, or migrating to server-side tagging. These carry real risk of breaking working tracking while you repair the broken parts, and the cost of getting it wrong is another quarter of data nobody trusts.

The honest test is this: if you cannot confidently explain what a change will do to your historical data, get help before you publish it.

Where this leaves you

Bad data is not a reporting inconvenience. It is a budget problem. Every campaign decision, every optimization, every smart bidding algorithm learning from your conversions is downstream of whether these events are correct.

Run the checks above. Give yourself an afternoon. If your conversion counts reconcile, your events are named consistently, and your tags fire only where they should, you are in better shape than most.

If they do not, you now know exactly where to start.

FAQ

How do I know if my GA4 is set up correctly?

Compare GA4 conversion counts against your CRM or ecommerce backend for a full month. Then check that each conversion fires once, not twice, using Realtime and GTM preview mode. Review your events list and remove anything nobody can explain. If your numbers reconcile within a reasonable margin and every event has a known trigger, your setup is probably sound.

Why does my GA4 data not match Google Ads?

Some difference is expected because the two use different attribution models and windows. Google Ads credits conversions to the click date, GA4 to the conversion date. Large gaps usually mean something else: duplicate tags firing on one platform, conversions defined differently in each, or missing deduplication so a single purchase is counted more than once.

What causes duplicate conversions in GA4?

The most common cause is having the same tag deployed twice, once hardcoded in the site template and once through Google Tag Manager. Others include triggers that fire on both a click and the resulting page load, thank you pages users can refresh, and transaction IDs that are missing or not unique, which prevents GA4 from recognizing repeat events.

How often should I audit my GA4 setup?

A full review makes sense once a year, and any time your site changes materially. A redesign, a new checkout flow, a CMS migration, or a new payment processor can all silently break tracking. A shorter monthly check comparing conversion counts against your source of truth will surface most problems long before an annual audit would.

Can I fix GA4 tracking issues myself?

Settings level fixes are reasonable to handle in house: data retention, internal traffic filters, unwanted referrals, and unmarking events that should not be conversions. Container level work is riskier. Rebuilding conversion tracking or resolving duplication can break functioning tags while repairing broken ones, so bring in help when you cannot predict what a change will do.