These prompts help product teams use AI to analyze funnel data — from diagnosing drop-offs to measuring the impact of changes and planning tracking for new flows. Replace [bracketed placeholders] with your specifics.
Prompt 1: Diagnose funnel drop-offs and prioritize fixes
I have funnel data for [product name]'s [flow name] flow.
Funnel steps and metrics:
[Paste: step_name, users_entered, users_completed, conversion_rate, median_time_seconds]
Segment data (if available):
[Paste: segment (e.g. mobile/desktop), step_name, conversion_rate]
Context: [product description, recent changes, known issues]
Please:
1. Rank the steps by absolute drop-off (users lost) and identify the top 3 priority areas
2. For each priority drop-off, list 3-5 probable causes based on common UX friction patterns
3. Flag any segment where conversion is significantly worse than the overall average
4. Recommend a specific qualitative follow-up method for each major drop-off
5. Estimate the business impact of fixing the top drop-off (if improving conversion by 10-20% at that step, how many more users complete the funnel?)
6. Prioritize the fixes by (impact × feasibility)Prompt 2: Compare funnel performance before and after a change
I made a change to [step name] in our [flow name] funnel. Here is the funnel data for both periods:
Before (dates: [start-end]):
[Paste: step_name, users_entered, conversion_rate]
After (dates: [start-end]):
[Paste: step_name, users_entered, conversion_rate]
Context: [what was changed and why, any external factors that might affect the comparison]
Please:
1. Calculate the change in conversion rate at each step
2. Run a significance test (z-test for proportions) for the changed step — is the improvement/decline statistically significant?
3. Check whether the change affected downstream steps
4. Check for unintended effects (did any other step's conversion worsen?)
5. Provide a concise summary: did the change work, and what should the team do next?Prompt 3: Design a funnel tracking plan for a new flow
We are building a new [flow name] flow with these planned steps:
[List the user-facing steps of the flow]
Product context: [what the product does, who the users are, what the business goal of this flow is]
Please create a funnel tracking plan that includes:
1. For each step: the event name (following [naming_convention] format), when it fires, and key properties to capture
2. The recommended conversion window for this funnel type
3. Key segments to analyze (device, user type, traffic source, etc.)
4. Alert thresholds: what conversion rate at each step should trigger investigation?
5. Companion metrics to track alongside the funnel (time between steps, error rates, help page visits)