ArGA4 lead tracking: lessons from the 2022 transition
The archive date follows the period covered. This retrospective was added on November 3, 2026.
A website can generate plenty of clicks while delivering very few useful enquiries. That is why GA4 lead tracking should start with a business definition, before anyone adds another tag.
The 2022 milestone
On March 16, 2022, Google announced that standard Universal Analytics properties would stop processing new data on July 1, 2023. Its GA4 transition announcement put an event-based measurement model at the centre of the move. This is a retrospective on that change, with implementation advice reviewed in 2026.
Define the moment a lead actually exists
For a service business, three actions often get mixed together: opening the contact page, pressing Submit and having an enquiry accepted by the application. We would measure these separately. A click shows intent; it does not prove that a valid message reached the backend.
Consider a form that rejects an invalid email address. If tracking fires on the button click, the analytics report can count a lead even though the visitor still sees an error. An accepted submission is a stronger trigger. It still does not prove email delivery, sales qualification or a completed sale; those are separate stages.
Google documents generate_lead for lead generation. Our implementation recommendation is to emit it after the application confirms acceptance, then test that a double click or repeated callback does not produce duplicate events.
A small measurement plan beats a long event list
Before implementation, agree on a short event dictionary. For each event, write down the trigger, the owner, the reporting question and the conditions that should prevent it from firing.
- Contact intent: an email link, phone link or contact-page visit. Useful for understanding the journey.
- Accepted enquiry: a validated form submission accepted by the backend. Useful for measuring website lead generation.
- Qualified opportunity: a later sales stage recorded in the CRM. Useful for assessing lead quality when a separate integration is in scope.
Keep enquiry names, email addresses and message text out of event parameters. Service categories and form locations can usually answer the website question without copying personal form content into analytics.
Test the paths that can distort the report
Our suggested acceptance checks include a successful submission, invalid input, a server error, repeated clicks, a page reload after success and the site’s consent states. Use a test environment or a clearly identified test lead so the QA exercise does not confuse sales reporting.
Google’s event setup guide describes checking events in Realtime and DebugView. That confirms what Analytics receives; application logs and the lead destination answer different questions. A complete handover should identify which layer each check covers.
Use the result to choose the next experiment
If many visitors open the form but few enquiries are accepted, inspect the form experience before buying more traffic. If accepted enquiries look healthy but qualified opportunities remain low, investigate message fit and targeting with the sales team. Those are hypotheses to test, not conclusions the event count can establish alone.
At ArAgile, we can connect front-end implementation, API work and experimentation and CRO development. A sensible first deliverable is a measurement specification with tested triggers, rather than a dashboard full of unexplained numbers.
Plan your analytics and experimentation implementation
Tell us about your existing setup, the process you want to improve and the constraints that matter.