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

# Build lists with chat

> Find and refine companies and people in chat.

## Overview

Use chat to find companies and people that match your criteria. You can begin
with a description of your target buyers, an existing Unify list, or a file of
accounts you want to investigate. Chat organizes the results in a table that
you can refine before saving a list or preparing outreach.

Your criteria can combine company and person information with research and
connected CRM data. For example, find accounts in a particular industry, narrow
them to companies with an open opportunity, and identify people in the roles
you want to reach.

## Build your list

**1. Define who you want to find**

Open [chat](https://app.unifygtm.com/dashboard/chat). Describe the companies
or people you want, how many results you need, and the evidence that should
qualify them.

Include the criteria that matter to your outreach, such as company size,
location, industry, role, or a recent business development. If you want to
start with existing data, type `@` to select a list or attach a CSV.

For example:

> From my target account list, find companies with 100–500 employees in
> the United States. Add their industry and account owner to a table.

**2. Refine the table**

Review the results and ask chat to make specific changes. You can add
columns, enrich missing information, research each row, or apply a
qualification rubric.

For example: “Add a column explaining whether each company sells to
other businesses, with a source for the answer.”

Use the table's column controls to sort and filter the displayed results.
When an enriched cell has supporting evidence, double-click it to review
the reasoning and references. Use that evidence to assess whether the
result meets your criteria.

Chat showing three saved accounts in a table with domains, headcounts, and owners.

**3. Save the results**

Click **Save as Unify list** to save every record in the table, or ask chat
to save the companies or people to a named list. Specify whether you want
a new list or an existing one.

To save only part of a table, include the criteria in your request. For
example: “Save the companies with at least 100 employees to a new list
named Mid-market planning accounts.”

Chat summarizes which records were added and identifies any records it
could not save. Open the resulting list to review its members. Companies
need a website domain and people need an email address to be saved from
new prospect data.

Chat confirming a new list containing the two accounts with at least 100 employees.

## Use your results

Once you have reviewed the results, you can prepare outreach from the same
conversation. Click **Add to sequence** or ask chat to prepare a sequence for
the people you selected. For personalized drafts and their approval workflow,
see [Agentic sequences](https://docs.unifygtm.com/reference/sequences/personalization/agentic-sequences.md).

To work with the table outside Unify, use its download control to export a CSV.
The export follows the table's current filters, search, and sorting. A single
export includes up to 50,000 rows; narrow the table if you need to export a
larger result in smaller groups.

You can return to the conversation to keep refining the table. To use it in
another conversation, type `@` and select it from **Chat tables**.

## Troubleshooting

> [accordion] Why is a result missing information?
>
> Research and enrichment depend on the information available for that
> company or person. Ask chat to investigate the missing value and show its
> sources. Review the result before relying on it for qualification or
> personalization.

> [accordion] Why could chat save only part of the list?
>
> Review chat's result for invalid or failed records. Check that each company
> has a usable website domain or each person has an email address, then ask
> chat to retry the affected records. Duplicate inputs and records already
> in the list are counted separately from newly added members.


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