Skip to main content

​
Funnels

Funnels measure how people progress through a sequence of steps toward a defined outcome. They help answer not only how many people complete a workflow, but where the largest losses occur. Typical funnels include landing page → signup → activation, pricing page → checkout → purchase, or onboarding started → setup completed → first meaningful action.

​
Design a useful funnel

Before creating a funnel, define:
  1. The population: Which visitors or users should be included?
  2. The entry point: What event or step starts the flow?
  3. The completion point: What counts as a successful outcome?
  4. The sequence: Which intermediate steps matter for the question?
  5. The time window: Over what period should users be allowed to progress?
Use events that represent meaningful progress. A click on a “Create account” button is not the same as a successfully created account.

​
Build and read a funnel

1

Choose the workflow

Select one high-value flow. Start with a short sequence so the result is easy to interpret.
2

Define each step

Use the available Trfyx event or page conditions to define the entry point, intermediate steps, and completion event. Keep definitions consistent when comparing time periods.
3

Set relevant filters and time range

Limit the report to the audience and period relevant to your question. For comparisons, avoid changing several conditions at once.
4

Find the largest drop-off

Identify the stage where the biggest number or proportion of users fail to continue. Check whether the result is stable across different periods or segments before prioritizing it.
5

Investigate the experience

Use user journeys to explore alternate paths and session replay to look for friction in the relevant step.
6

Validate a change

Make one targeted improvement and compare the funnel after enough data has accumulated. Account for traffic mix and any tracking changes during the comparison.

​
Interpreting drop-off

A drop-off is a signal to investigate, not a complete explanation. Users may leave for many reasons: confusing copy, a technical error, an expected exit, poor traffic quality, or a step that is not required for every user. Ask these questions before changing the product:
  • Is the step instrumented correctly?
  • Is the next step required for every user in this funnel?
  • Does the drop-off differ by device, source, or audience?
  • Do recordings or support reports reveal a consistent obstacle?
  • Did a release or campaign change the mix of visitors?

​
Common funnel mistakes

Too many steps. Begin with the core flow and add detail only when it helps diagnose a question. Inconsistent events. Renamed or missing events can make historical comparisons unreliable. Confusing sequence with causality. A step that correlates with conversion does not necessarily cause it. Comparing different audiences. A funnel for high-intent returning visitors may behave differently from one for all first-time visitors. Start with goals to define the successful outcome, use analytics overview to establish the baseline, and inspect session replay when a stage needs deeper investigation.