# Find where deals leak out of my sales funnel

A stage-by-stage conversion and velocity analysis that pinpoints where deals stall or drop, and why, with targeted fixes.

- Topic: Data & Analysis
- Framework: Few-Shot
- Level: Standard
- Tags: Sales, Analyze, Fill-in Template
- Source: https://www.thinkwithprompt.com/prompts/find-where-deals-leak-out-of-my-sales-funnel

## Prompt

```text
Help me find where my sales funnel is leaking.

**Funnel data:** [paste counts and values by stage for a period: leads, meetings, opportunities by stage, closed-won, closed-lost, with dates if possible]
**Average days in each stage:** [if available]
**Segments to compare:** [e.g. region, source, segment, rep, product]
**Comparison period:** [e.g. same period last year, prior quarter]
**Our stage definitions:** [paste, or describe briefly]

Steps:
1. Calculate stage-to-stage conversion rates and time in stage for the period and the comparison period.
2. Find the 2-3 stages with the biggest drop in conversion or biggest increase in time. Show the change.
3. Break those stages down by the segments I listed to see if the leak is everywhere or concentrated.
4. Calculate sales velocity (opportunities x win rate x deal size / cycle length) and which component changed most.
5. Offer likely causes for each leak and how to confirm each (e.g. listen to calls at that stage, check loss reasons, check if stage exit criteria are followed).

Return:
- A funnel table: Stage | Conversion now | Conversion before | Days in stage now | Before.
- The top leaks, each with evidence and likely causes.
- A velocity breakdown.
- 3 fixes ranked by impact and effort, and what metric would show each is working.

Don't overstate conclusions from small numbers; flag stages with low counts.
```

## What to fill in

- if available
- e.g. region, source, segment, rep, product
- e.g. same period last year, prior quarter
- paste, or describe briefly

Free to copy and use. From ThinkWithPrompt (https://www.thinkwithprompt.com/), a library of 2,663 curated AI prompts.
