Start with what your site is meant to measure
Before opening an analytics report, write down the actions you expect it to record. For a small content site, these might include page views, visits to important guides, search within the site, newsletter sign-ups, or clicks on a contact button. Separate actions that indicate a useful outcome from actions that are merely easy to count. A high number of button clicks, for example, does not prove that people completed the form behind the button.
Next, identify where tracking is installed and which analytics property should receive the data. It may be added directly to the site, through a tag manager, or by a website platform. If more than one person has worked on the setup, check for overlapping installations. Note the pages and devices you will test, including templates that behave differently, such as articles, forms, and confirmation pages. This short inventory gives you a clear standard for deciding whether tracking works.
Run a controlled visit and look for your activity
Open your analytics tool’s real-time or debugging view, then visit the site in a separate browser window. Navigate through several pages and perform one action you expect to be tracked. Use pages or an action you can easily recognize in the report. If possible, run the test when few other people are likely to be on the site; otherwise, treat what you see as suggestive rather than definitive.
Allow for a short delay and check whether the expected page views and events appear in the correct property. Confirm more than their presence: look at the page path, event name, and any useful details attached to the event. Repeat the test on a mobile device or browser if those experiences differ. If nothing appears, check that you have the right property selected, the tracking code loads on the tested page, and your own traffic has not been excluded by a filter.
Check for missing, duplicate, and misleading events
A working tag can still produce unreliable reports. Refresh a page once and inspect the result. If one visit appears to create multiple page views, look for duplicate tracking installed through both the website and a tag manager. Then test the same action twice, deliberately. An event that fires before a form is successfully submitted may be measuring intent to submit rather than completion. Where a confirmation page exists, test that it appears only after the action succeeds.
Follow a typical visitor journey across the site. Check whether navigation, redirects, or a move to another domain unexpectedly start a new visit or obscure the original traffic source. Pay particular attention to third-party booking, payment, or form tools if they are part of an important outcome. You may not be able to track every step, but you should know which steps your reports do and do not cover. Name events consistently so that the same action is not split across several labels.
Account for consent and browser differences
Tracking tests should reflect the choices visitors can make. If your site uses a consent banner, test both an accepted and a declined choice. Verify that the tracking behavior matches your intended consent setup; do not treat missing data from someone who declined tracking as a technical fault to work around. Privacy settings, content blockers, and browser behavior can also affect what appears in analytics.
This is why analytics totals should not be interpreted as a perfect count of every person who used the site. Compare patterns over time with that limitation in mind. If a test behaves differently across browsers or devices, record the difference and investigate it before drawing conclusions from the affected report.
Keep a simple tracking check log
Create a small record of each check: the page tested, the action taken, the expected result, the observed result, the browser or device used, and whether consent was given. Include the analytics property and any relevant filters. If you find a problem, fix one likely cause at a time and repeat the same test. This makes it easier to tell whether a change solved the issue or introduced another one.
Recheck tracking after changes to templates, forms, consent settings, tag manager rules, or the analytics configuration. Before using a metric in a report, ask whether it still represents the action its label suggests. A modest set of clearly defined, regularly tested measurements is more useful than a crowded dashboard whose numbers no one has verified.
