# Conversion funnel analysis (/data--funnel-analysis)

/data--funnel-analysis is a Claude Code skill in the Data section. Builds conversion funnel SQL on your real event data, with drop-off per step and per segment, plus an insights template for next steps.

- Web version: https://skills.sgomez.dev/en/s/data--funnel-analysis
- Section: [Data](https://skills.sgomez.dev/en/data.md)
- Author: Santiago Gómez de la Torre
- License: MIT
- Source: https://github.com/sgomez-dev/claude-skills/blob/main/skills/data/funnel-analysis.md
- Updated 10 Jul 2026

## Use it when

- You want to know which onboarding or checkout step loses users
- You need to compare conversion by channel, device or plan
- You have tracked events and want to prioritize what to fix first

## Not for

- Measuring return behavior over time, which is what cohort analysis is for
- Connecting to your warehouse and running the queries for you

## What you get

A funnel_name.sql file with the funnel definition and a funnel_name.md with a pre-structured insights template.

## How to ask for it

- `/data--funnel-analysis conversion funnel for onboarding`
- `/data--funnel-analysis SQL to find where users drop off at checkout`
- `/data--funnel-analysis analyze the funnel from free trial to paid`

## Install

macOS · Linux:

```
curl -fsSL https://raw.githubusercontent.com/sgomez-dev/claude-skills/main/install.sh | bash
```

Windows:

```
irm https://raw.githubusercontent.com/sgomez-dev/claude-skills/main/install.ps1 | iex
```

Claude Code plugin:

```
/plugin marketplace add sgomez-dev/claude-skills
/plugin install data-skills@claude-skills-collection
```

## Permissions

- Reads: `**/*.sql`, `**/models/**`, `**/migrations/**`, `**/*.yml`, `**/schema*`, `**/*.prisma`, `**/*.py`
- Writes: `funnel_*.sql`, `funnel_*.md`
- Runs: —
- Network: No
- Destructive: No

## Author's description

Build a conversion funnel: step definitions, drop-off SQL, insights template

## Questions about this skill

### Does it run the queries against my database?

No. It only writes the SQL and the documents; you run the queries yourself against your warehouse.

### What if an event for one step is not tracked?

It flags the gap instead of quietly substituting a similar event. For defining it, it points you to the event tracking plan skill.

### Why rank by users lost instead of percentage?

A small percentage leak on high volume can lose more people than a large one on low volume.


- [How we review this](https://skills.sgomez.dev/en/methodology.md)