Define the event sequence
Give each event a stable name, trigger, owner, and required properties. Do not infer a result view from completion if rendering can fail.
| Event | Definition | Question answered |
|---|---|---|
| Eligible view | Quiz was visible and usable | Did the opportunity exist? |
| Start | First answer or explicit start action | Did the promise earn engagement? |
| Progress | A defined question milestone | Where does friction appear? |
| Completion | All required quiz questions finished | Did participants reach a result? |
| Result viewed | Result rendered successfully | Was value delivered? |
| Contact submitted | Requested details accepted | Was the exchange compelling? |
| Next step | Result-specific action selected | Did the result create intent? |
| Qualified outcome | A business-defined quality event | Did the journey create value? |
Calculate rates with explicit denominators
Always label the denominator. Two teams can report different “quiz conversion rates” from the same data because one uses views and another uses starts.
- Start rate = starts divided by eligible views
- Completion rate = completions divided by starts
- Result delivery rate = results viewed divided by completions
- Lead capture rate = contact submissions divided by result views or starts, state which
- Next-step rate = next-step actions divided by result views
- Qualified yield = qualified outcomes divided by eligible views
Break results down before changing the quiz
Segment by traffic source, campaign, device, new versus returning visitor, quiz result, and contact-gate position. A single sitewide average can hide a strong high-intent segment and a weak low-intent campaign.
Use the metric pattern to diagnose the problem
A metric points to an investigation. It does not prove a cause without a controlled change or additional evidence.
| Pattern | Likely investigation |
|---|---|
| Low start, normal completion | Promise, placement, traffic match, or load experience |
| High start, early drop-off | First questions, expectation mismatch, or mobile usability |
| High completion, low next step | Result usefulness, trust, or call-to-action fit |
| High lead volume, low qualification | Traffic targeting, scoring rules, or overly broad promise |
| One result dominates | Question balance, routing rules, or audience mismatch |
Use a weekly decision report
Report one traffic metric, one experience metric, one lead metric, one quality metric, and one experiment decision. This keeps the team from celebrating higher completions while qualified yield declines.
Evidence
How to reproduce the method
The event dictionary and formulas make the framework reproducible across analytics tools. Validate each event with a test submission, then reconcile event counts against raw submissions for a defined period.
Limitation
Where this conclusion stops
This framework does not provide universal benchmarks. Rates depend on traffic intent, placement, audience, offer, question burden, device, privacy settings, and the definition of a qualified outcome.
Sources and verification
What this guide relies on
Sources and method checked August 7, 2026. External standards are linked to their primary publishers.