Back to Prompt Library
Data Analysis
funnel analysisconversionproduct analyticsCROSQL

Funnel Analysis Plan

A complete funnel analysis plan with stage definitions, drop-off diagnostics, query skeletons, and experiment hypotheses.

Best For

  • A growth PM diagnosing where users drop off between free trial signup and paid conversion before proposing an experiment
  • An e-commerce analyst comparing checkout funnel conversion by device type to prioritize a mobile-first fix
  • A data analyst writing a reusable GA4 or Amplitude query skeleton so funnel reviews don't require rebuilding the report each quarter

Prompt Template

Create a funnel analysis plan for [FUNNEL_NAME, e.g., the free-to-paid conversion funnel / the e-commerce checkout funnel] at [COMPANY_NAME]. The funnel covers the journey from [FUNNEL_START] to [FUNNEL_END]. Define the analysis as follows: (1) Funnel stages, list each stage in order with: stage name, the specific user action that defines entry into the stage, and the data event or field that captures it in [ANALYTICS_PLATFORM, e.g., Mixpanel, Amplitude, GA4, SQL], (2) Conversion rate baseline, current conversion rate at each step: [CURRENT_RATES] and the industry benchmark if known, (3) Drop-off analysis, for the 2 highest drop-off stages, specify: what data to pull to diagnose why users drop off (session recordings, heatmaps, exit surveys, error logs), (4) Segmentation cuts, which dimensions to slice the funnel by: [DIMENSIONS, e.g., acquisition channel, device type, plan type, geography], (5) SQL or analytics query skeleton, provide a query skeleton for calculating stage-by-stage conversion in [DATA_WAREHOUSE], (6) Experiment hypotheses, for each major drop-off point, propose 1 specific experiment to improve conversion with a hypothesis statement, (7) Reporting cadence, how often this funnel should be reviewed and by whom.

Pro Tip

Name the specific event or field that defines entry into each stage before writing any query, since funnel stages defined loosely in words almost never match how the analytics platform actually logs the event.

Example Output

A sample of what this prompt produces once you fill in the placeholders.

Funnel: trial signup to paid conversion. Stage 2, 'activated,' defined as completing onboarding step 3 within 48 hours, current conversion 38% versus an industry benchmark near 45%. Highest drop-off: activation to first paid invite, at 22%. Experiment hypothesis: adding an in-app checklist will lift activation by reducing time-to-first-value.

How to use this prompt

  1. Copy the prompt template using the button above.
  2. Paste it into your preferred AI assistant (ChatGPT, Claude, Gemini, etc.).
  3. Replace all bracketed placeholders like [TOPIC] with your specific details.
  4. Send the prompt and refine the output as needed.