A direct, step-by-step process for confirming your Google Analytics tracking is actually accurate, not just installed, covering duplicate tags, conversion event verification, and cross-checking against Search Console.
Installing Google Analytics and having it actually work are two different things. A tracking code can be present on a page, show no errors, and still be silently recording nothing useful, or worse, recording data that looks plausible but is wrong in a way that won't be obvious for months. This is a direct, step-by-step process for confirming your tracking is genuinely accurate, not just installed.
Most setup guides end at installation: paste the code, see the property exist, move on. The problem is that a surprising number of common mistakes, a tag installed on some pages but not others, a duplicate tag firing twice and inflating numbers, an ecommerce event that never actually fires, don't produce any visible error. The dashboard looks normal. The mistake only becomes obvious later, usually when a number looks wrong enough to investigate, by which point weeks or months of decisions may have already been made on bad data.
Open your GA4 property and go to Reports, then Realtime. In a separate browser tab or an incognito window, visit your own website. Within a few seconds, you should see yourself appear as an active user in the Realtime report. If nothing appears after a full minute, the tracking code either isn't installed correctly or isn't firing on the page you tested. Check a few different pages on your site individually, not just the homepage, since it's common for a tag to be correctly installed sitewide through a template but missing on one specific page type that uses a different template.
Duplicate tags are one of the most common and least visible tracking problems, because they don't cause an error, they simply inflate every number by counting each real visit twice or more. In your browser, open Developer Tools, go to the Network tab, and reload your page while filtering for requests containing “collect” or “g/collect”. If you see multiple nearly identical requests firing for a single page load, that's a strong signal the tracking code is installed more than once, commonly because it was added directly to the site's code and also through a plugin, page builder integration, or a previous developer's leftover installation.
Enhanced Measurement automatically tracks events like scrolls, outbound clicks, and file downloads, but automatic doesn't always mean guaranteed. In GA4, go to Admin, Data Streams, select your web stream, and review which Enhanced Measurement events are toggled on. Then go to Reports, Realtime, and scroll to the event count section while browsing your own site, scroll down a page, click an outbound link, and confirm each corresponding event appears in real time.
This is the step that matters most for any business actually trying to measure results, not just traffic. In GA4, go to Admin, Events, and confirm the specific actions that represent real business value, a form submission, a purchase, a phone click, a scheduled call, are marked as key events. If your key event doesn't yet exist as a trackable event at all, it needs to be configured before it can be marked as a conversion.
Test each key event manually: submit a real test form, click a real phone number link, walk through a real test purchase if you run ecommerce. Confirm each action appears correctly in the Realtime report under DebugView, which shows a more detailed, developer-oriented view of exactly which events and parameters are firing as you interact with the site.
If Search Console is linked to your GA4 property, compare organic session counts between the two tools for the same date range. They won't match exactly, since the two tools measure slightly different things, but a wildly large gap, sessions in one tool dwarfing the other by several multiples, is a signal worth investigating rather than ignoring.
Confirm your own team's visits are actually being excluded, not just configured to be excluded. Check the Tech report under Acquisition, and look for an unusually high share of sessions from your own office location, VPN provider, or a small handful of specific devices that are clearly internal rather than genuine customer traffic.
A quick check after any website redesign, platform migration, or major page template change is worth doing every time, since these are the events most likely to silently break a previously working tag. Beyond that, a full verification pass once or twice a year catches drift that accumulates gradually.
The Realtime report shows a general count of active users and top-level events. DebugView shows a much more granular, event-by-event stream specifically for a device you've manually put into debug mode, including individual event parameters, which makes it far more useful for confirming a specific conversion event is firing with the correct data attached.
Yes. Plausible-looking numbers are exactly what a duplicate tag, a partially broken conversion event, or an internal-traffic contamination problem produces. Reasonable-looking data and accurate data are not the same thing, and the only way to tell the difference is to actually run the verification steps rather than trust that things look fine.
Once your core tracking is confirmed accurate, the next layer worth adding is Microsoft Clarity, which shows you the actual behavior behind the numbers through heatmaps and session recordings. This connects directly to our Google Analytics 4 and Google Search Console setup guides. If you'd rather have your full tracking stack audited and verified by someone who does this daily, that's exactly what our paid media audit covers, and it's part of every free Growth Gap Analysis.
How We Work
