Check the live page first
Make sure the change is published, not just saved in an editor or preview. Open the public page source and find the tracking script. Confirm that the website ID matches the website you are viewing in the dashboard.
If your site uses a page cache or CDN, check that the version delivered to a normal visitor contains the tag. Visit both the home page and a second template, such as an article or product page. Missing shared code often explains why only part of a website is visible in analytics.
- Use the generated tag for this website.
- Install one copy per page.
- Publish your changes and refresh the relevant cache.
- Check the shared header and any separate page templates.
Look for the collection request
Open your browser’s developer tools, select Network, and reload the page. The script.js download should succeed. The tag should then send a pageview to the collection endpoint, normally /api/send on the tracking service.
A failed script download points to the script address, availability or browser blocking. A failed collection request can point to a content security policy, network problem, invalid payload or blocked endpoint. Read the status and the browser’s error message before changing settings. Do not work around a visitor’s privacy choices.
Check consent and the report settings
If the website waits for consent before loading analytics, a visit that has not given that consent may correctly produce no analytics request. Other common exclusions include your own logged-in browser, particular domains or a browser setting configured to disable tracking.
In the dashboard, clear optional filters, select a period that includes the test, and check the reporting timezone. A country exclusion can hide your own test traffic. Use Realtime and the installation panel to check for recent events before interpreting a longer report.
Keep a small, repeatable test
Write down the page URL, the time and timezone, the browser, and the action you performed. Record the script and collection response status without copying personal information or credentials. Repeat one page load and one navigation.
A consistent test makes it much easier to distinguish a delivery problem from a counting difference. Once the requests are arriving, compare pageviews first and then investigate visitor or visit grouping if needed.