> ## Documentation Index
> Fetch the complete documentation index at: https://docs.rippit.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Taxonomy Builder

> Contact reasons from what customers say, not a preset dropdown.

**B2B** · Customer experience · For Support ops and VoC teams · Also for [B2C companies](https://docs.rippit.com/use-cases/taxonomy-builder-b2c.md)

Taxonomy Builder builds your contact reasons from the customer’s own words, not your team’s best guess, across every ticket, chat and AI agent conversation. Every conversation gets the right reason, mis-tags are flagged, and new issues are proposed while they are emerging. The drivers you report hold up when product and leadership dig in.

## What it reads

Help desk tickets, chats and email, in-app chats, shared Slack channels, your AI agent's conversations, your knowledge base, and your current tags and fields, plus any other source you integrate or bring in through bulk ingest.

## How it works

1. **Ingest.** Every conversation, plus the tags you have today.
2. **Analyze.** The real reason on every conversation.
3. **Act.** A taxonomy that stays current.

## What you get

Contact reasons, fixed mis-tags and new issues flagged as they emerge.

* **Contact-reason taxonomy** (Notion or Google Drive): 9 categories, 38 reasons, each with a definition and example conversations.
* **Help desk fields** (Your help desk, through MCP): 1,870 corrected tags, ready to write back through a custom MCP connection.
* **Any tool you connect through MCP:** Send the weekly top-reasons report to Slack, or open a Linear issue for the team that owns a fast-growing reason.

## Build it

Tell Rippit chat to "set up contact reasons built from what customers actually said, with mis-tags fixed," or ask your own AI assistant with Rippit connected. Rippit drafts it as a [pipeline](https://docs.rippit.com/pipelines/overview.md), tests it on a sample, and puts it live when you confirm. You can change what it reads, how it scores, who it notifies, and how often it runs at any time.

## Questions

> [accordion] Do we have to start from scratch?
>
> No. It starts from your current taxonomy, keeps the categories that hold up, and suggests merges, splits and new reasons where the conversations say so. You approve every change to the structure.

> [accordion] Will it change our help desk tags on its own?
>
> Only through a write connection you approve. Rippit corrects the reason in its own table first. To change tags in your help desk, an admin connects it through a custom MCP connection and turns on the write tool, or your team reviews a weekly list of changes.

> [accordion] When does it propose a new category?
>
> When a theme that doesn't fit your taxonomy crosses the volume you set, like 200 conversations in a week. You get a proposed name, a definition and example conversations to approve.

> [accordion] Where do the results go?
>
> Corrected reasons live in Rippit and can go back to your help desk through a custom MCP connection, or as internal notes in Intercom. The taxonomy doc lives in Notion or Google Drive, and a weekly summary can go to Slack or email. It runs every Monday by default, and you can change the schedule any time.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.