Create intermediate Works with: Any AI

Funnel Analysis and Drop-Off Diagnosis

Table output Has example

Find where users leave.

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

Analyze a funnel. Steps and counts: [FUNNEL_COUNTS]. Segments (device, source, new vs returning): [SEGMENTS]. Recent changes: [RECENT_CHANGES].

Compute step conversion (count at step / count at previous step) and cumulative conversion, with arithmetic, and identify the biggest drop. Compare segments. For the biggest drop-offs, generate hypotheses (usability, performance, trust, pricing surprise, mismatch of expectations), each with data to check (session recordings, error logs, surveys) and a quick test.

Produce a table: Step | Users | Step conversion | Cumulative | Drop-off | Hypotheses | Evidence to gather. Rank opportunities by potential gain: if the step improved by X points, how many more completions (computed). Add warnings about sample size, bots, tracking gaps, and seasonality. Provide a plan for the next two experiments.

List the three most valuable fixes by estimated additional completions per month, using the arithmetic, and propose how to confirm each cause with session recordings, error logs, or five user tests before building anything.

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.

Funnel: Visit 10,000; Signup 1,800 (18.0 percent); Create project 1,080 (60.0 percent of signups); Invite teammate 432 (40.0 percent); Paid 86 (19.9 percent).
Cumulative: signup 18.0; project 10.8; invite 4.3; paid 0.86.
Biggest drop by count: visit to signup (8,200 lost); by rate: invite to paid.
Mobile: signup rate 11 percent versus desktop 23 percent.
Hypotheses: mobile form is long; pricing surprise before paid.
Value: if mobile reached 17 percent (+6 points on 5,000 mobile visits) = +300 signups; at 4.8 percent paid-from-signup that is 14 more customers.
Experiments: shorten mobile form; show pricing earlier. Check bot traffic and tracking gaps first.

• funnel