GA4 conversion tracking not working? Trace one test lead
Diagnose a missing or duplicate GA4 key event by checking the form backend, GTM trigger, and GA4 destination.
Start with one test lead and note the timestamp, page URL, and browser. GA4 calls its important Analytics events key events. If that event also feeds a Google Ads conversion, check the Ads destination separately.
Pick the branch that matches your test
- No lead in the backend. Start with the form. Check the response status and success state, then reproduce on the affected browser. A click or success message alone can appear even when the lead was never saved.
- Lead saved, no event in GA4. Trace the success trigger, tag, and destination. Check consent and the measurement ID. Capture one test timestamp so you can follow the same attempt through each tool.
- One lead, two events. Compare the firing sequence in Tag Assistant. Look for a click trigger and a success trigger both sending the same event, two implementations sending it, or a confirmation-page reload repeating it.
- Event visible, totals still differ. Compare equivalent windows and event definitions, including time zone, filters, and deduplication. Check the Ads conversion separately if it has its own attribution and counting settings.
Follow one successful submission
- Check the source of truth. Verify the test lead arrived in your CRM or form backend. If it didn’t, inspect the form request and the error shown to the visitor.
- Check the trigger. Open GTM Preview and inspect the tag firing sequence with Tag Assistant. Submit successfully once, then try an invalid input and reload the confirmation page. The success tag should fire once for the successful submission. Tag Assistant troubleshooting
- Check the destination. Verify the measurement ID and use GA4 DebugView for the test session. Confirm the event name matches the one configured as a key event. Google tag troubleshooting · DebugView
- Check consent paths. Repeat with the relevant consent choices in a clean browser. Compare actual requests with your documented expectations for each state and record the result.
If the CRM has more records than GA4, compare the same date window and definitions before attributing the gap to broken tracking. Consent, browser blocking, duplicates, and reporting filters can all change which events you see.
Record the backend result, trigger count, destination, consent choice, and observed event count. Rerun the test after changing a trigger so you can compare the same path.
Use the marketing site release checklist to put this test alongside your mobile and indexing checks.