Afield

Afield resources

Define a content conversion event before reporting the rate

A filled event definition and test matrix for a lead form, with a 40-click funnel showing which conversion rates you can calculate.

In this article

For a page offering a sample review, a button click, an accepted request and a relevant sales inquiry are three different outcomes. Define each one before publishing a conversion rate.

Each step has its own observable completion condition.
Each step has its own observable completion condition.

Write the event contract

Demo: an article invites readers to submit a sample file for review. This is a proposed measurement contract, not a claim about Afield’s analytics implementation.

ItemDefinition
Event namesample_review_submitted
TriggerThe application accepts and stores a valid request, then confirms success.
ExcludeOpening the form, clicking Submit, invalid input and a failed request.
CountOne accepted request per request identifier; repeat delivery of that identifier must not create a second business record.
ContextOffer name, originating article and form version. Keep the submitted file and personal contact details out of the analytics payload.
QualificationA separate sales-review field: relevant, irrelevant or awaiting review.

Implementing duplicate handling is part of the application’s measurement work; assigning an event name does not provide it. In GA4, a business-important event can be marked as a key event. That setting still depends on the underlying event being sent correctly.

Test the boundary, including failed attempts

Test actionBusiness records expectedSubmission events expected
Submit with a required field missing00
Submit valid data; storage succeeds11
Click twice while that request is pending1 for the same request1 for the same request
Refresh the confirmation pageNo additional recordNo additional submission
Request fails before acceptance00
Response is lost after acceptanceRead the stored request before retryingReconcile by request identifier; do not infer failure from the timeout

Check both the application record and the collected event. A thank-you screen alone cannot tell you whether the event fired twice.

Calculate only the rates your data supports

The demo campaign records 40 offer clicks, 28 form opens and 10 accepted requests. Sales has reviewed all 10 and classed 6 as relevant.

  • Accepted requests per offer click: 10 ÷ 40 = 25%.
  • Relevant share of reviewed requests: 6 ÷ 10 = 60%.
  • Visitor-to-request rate: unknown; this dataset contains no visitor denominator.

If sales has reviewed only 6 requests, report “4 awaiting review” separately instead of counting them as irrelevant. If form opens look healthy but accepted requests fall, inspect the specific form failure. For Google Search clicks versus site sessions, use the report reconciliation checklist.

Related reading

Review a report and choose the next investigation

Open an editable starting task. Add your details before running it.

Review campaign results