MVA / FIELD GUIDES / Conversion tracking
Two form events. One inquiry. Which number goes in the report?
Two events can describe one inquiry. Agree on the outcome before adding the numbers together, so a tracking detail does not become a reporting mistake.
The number needs a definition.
Suppose someone fills out your contact form. Two analytics events appear. The sales team has one new inquiry. Which number should go in the report?
If the metric is accepted inquiries, use a checked acceptance signal and agree on how repeats are handled. Do not add every event with a promising name. A submission attempt and a successful result describe different moments, even when they occur seconds apart.
- 01
The visitor sends the form
In this example, a submission event records the attempt.
Submission signal - 02
The website accepts it
A separate event records the agreed acceptance result.
Acceptance signal - 1
Accepted inquiry
Count the checked acceptance signal for this outcome, with agreed rules for repeats.
These invented event paths illustrate the distinction. Your form may behave differently, and receipt by your team is a separate check.
Follow what happened after Send.
Ask whoever owns the form and tracking what causes each event. A button click may occur before the form passes validation. A form submission may be followed by a rejected request. A success message may appear before a downstream system finishes delivery.
When form measurement is enabled, GA4 can collect events called form_start and form_submit. They describe form interaction and submission. Those labels alone do not establish that your business received an inquiry or that it became a qualified lead.
A success event needs scrutiny too. Have the implementation owner check a successful request, a rejected request, and whether repeating or refreshing the journey produces another event. Ask for the observed behavior, alongside the event definition.
Decide what the report should answer.
How many submission attempts?
A submission signal can help count attempts. Document what triggers it and whether an attempt can produce repeats.
How many inquiries were accepted?
Use a checked acceptance signal, with agreed handling of repeats. Confirm it stays silent when the request is rejected.
How many reached the team?
Compare with the receiving system, using an agreed date range and population.
How many became qualified leads?
That judgment usually happens later. It needs an agreed sales definition and appropriate records.
Those counts can all be useful. Give them distinct names so a dashboard does not quietly substitute one for another.
The receiving system helps. It does not settle everything.
Compare accepted requests with the records your team actually receives. Align the time zone, period, form, and treatment of duplicates. Record whether you are counting inquiries, people, or events.
Even a carefully defined analytics total can differ because collection coverage and system behavior differ. A matching total alone does not validate every journey or establish which channel deserves credit. Investigate the difference rather than forcing the numbers to agree.
Once the count is useful, channel credit needs its own review. Good lead counts can coexist with incomplete source data. The report should explain both.
Keep names, contact details, and message contents out of ordinary analytics event payloads. The handoff your team needs is a clear metric definition, checked behavior, and the limits of the comparison.