Create intermediate Works with: Any AI

Funnel Analysis Plan

Structured brief Has example

Find where users drop off.

The Open buttons send the prompt text as currently shown below to that site only when you click. Fill in the placeholders first.

Plan a funnel analysis. Funnel steps: [FUNNEL_STEPS]. Data available and events tracked: [DATA_EVENTS]. Questions: [QUESTIONS].

Define each step precisely, how events map to steps, time windows, and handling of users who skip or repeat steps. Provide a table of conversion at each step with placeholders and calculation formulas, and a plan to segment by device, source, new versus returning, and cohort. Identify biggest drop-offs and form hypotheses for each with the evidence needed. Check for tracking gaps and suggest qualitative research to explain why. Include how to present findings without overstating causes.

Add a list of ten questions to ask the data team about how events are defined and deduplicated, and a method for reconciling funnel numbers with finance or sales figures. Provide a template for presenting the findings, with the biggest opportunity first, and a plan to re-run the analysis after changes are released.

Also add a list of the segments most likely to behave differently and a hypothesis for each, to be tested rather than assumed.

Fill in before use

Optional: type your details and the prompt above updates. Nothing is saved or sent.

Replace each bracketed placeholder with your own details, then paste into Any AI or another assistant.

Steps: 1 Visit pricing page, 2 Start trial, 3 Add payment, 4 Activate.
Window: 7 days; users who skip steps counted at the highest step reached.
Conversion (placeholders): 100% -> 12% -> 60% of those -> 85% of those.
Biggest drop: Visit to Start trial (88% leave). Hypotheses: unclear value; wrong audience; price shock.
Segments: device, source, new vs returning.
Tracking gaps: event "trial_started" missing on mobile web.
Qualitative: 5 interviews with people who visited but did not start.
Present as "where we lose people", not why, until validated.

• metrics
• structured-brief