> For the complete documentation index, see [llms.txt](https://evidano.gitbook.io/ailyze-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://evidano.gitbook.io/ailyze-docs/analyze-data/categorize-responses.md).

# Categorize Responses

Sort thousands of short answers into consistent categories, count them, and export the coded data.

This page shows you how to sort many short answers into consistent categories, count them, and export the results.

At the end you will have every response labeled with themes and codes, plus frequency tables and a coded-data export.

## Before you start

* A spreadsheet (one XLSX or CSV) with a column of text responses — survey verbatims, reviews, or support tickets.

When you have hundreds or thousands of short answers, you want each one **sorted into categories** and **counted**. That's content analysis — sorting text into named categories and counting them — and Evidano automates the heavy lifting.

## Set it up

Use **Analyze xlsx/csv files**:

1. Choose **Analyze text responses in selected columns** and pick the column(s) to code.
2. Generate (or supply) a [codebook](/ailyze-docs/reference/glossary.md#codebook) of [themes](/ailyze-docs/reference/glossary.md#theme) and sub-themes.
3. With multiple columns, pick a **Codebook Mode**:
   * **Separate codebook per column** — each question is coded with its own scheme.
   * **Single merged codebook** — all selected columns use the same scheme (useful when several questions probe the same thing).

<figure><img src="https://1178311500-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fi4GBXPMUDxszIpUVOARU%2Fuploads%2FjP9aHxOnF0g8Wvl4iwRf%2FCategorize%20Responses%201.png?alt=media&amp;token=d0ce8c33-7ba9-486d-95c3-75c720c6415b" alt="Choosing columns and codebook mode for categorization"><figcaption><p>Pick the columns first, then decide whether they share one codebook or get one each.</p></figcaption></figure>

## What you get: the coded data table

The per-column results page includes **Data Coded with AI Themes & Sub-themes**, with three tabs:

* **AI Labels as List** — each response with its assigned codes.
* **AI Labels as Checklist (with code justifications)** — each response with a checklist of codes *and the AI's reason* for each — great for auditing quality.
* **Theme & Sub-theme Frequencies** — the counts, ready to chart.

You can **Download CSV** to pivot the coded data yourself. (On Free, only the first rows are shown; Pro reveals all.)

<figure><img src="https://1178311500-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fi4GBXPMUDxszIpUVOARU%2Fuploads%2FKXH48BAMGY9nHexbJTVj%2Fimage.png?alt=media&amp;token=89a23e37-2e5f-4034-8ded-20431834448c" alt="The coded data table with per-response labels"><figcaption><p>Every row keeps its original text next to the labels assigned to it, ready to filter or export.</p></figcaption></figure>

## Counting and visualizing

* **Frequency Bar Chart (Themes)** and **(Sub-themes)** show how often each category appears.
* Add a **Segment** to split counts by group, and use **Split legend by** for a second dimension.
* The **Synthesis** section describes the distribution in words.

For a walkthrough of every chart, see [How to read your results](/ailyze-docs/analyze-data/how-to-read-your-results.md).

## Keep coding consistent and accurate

* Write clear **descriptions** for each code so the AI applies them consistently.
* Use **Edit Codebook** to merge overlapping codes or split a code doing double duty — see [Build a codebook](/ailyze-docs/analyze-data/build-a-codebook.md).
* On the **Elaboration of Themes & Sub-themes** section, use **Find Similar Quotes** to catch responses that should share a code, and remove mis-assigned quotes.
* Run **Calculate AI Accuracy** *(Pro)* to gauge coding reliability against your codebook.

## Documents versus spreadsheets

* For **spreadsheets**, categorization is row-by-row — ideal for surveys, reviews, and tickets.
* For **documents**, themes and question columns group excerpts by code across all files, and choice columns put each whole document (or each row of your table) into categories you define, as described below.

## In document analysis

To categorize whole documents, add a choice column in the builder. The ideas behind the table are in [Design Your Analysis Table](/ailyze-docs/analyze-data/design-your-analysis-table.md).

1. In **You will receive a table of answers and metadata**, click **Add** › **Add column** (or the **+** after the last column), type a name, for example "Overall tone" or "Mentions privacy concerns", and press Enter.
2. Click the type under the name and choose **Single choice (pick one)**, **Multiple choice (pick all that apply)**, or **Yes / No**.
3. Open **Edit column**, choose **Specify my own options manually now**, and type one option per row (Enter adds a row); click **Add descriptions** if an option needs explaining. Tick **Use both AI-generated and manually entered options** to catch categories you did not list. A **Yes / No** column needs no options.

You should see the options listed under the column name in the table.

<figure><img src="https://1178311500-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fi4GBXPMUDxszIpUVOARU%2Fuploads%2Fgit-blob-b088203c241954f865bac165c7e5dee09fbdd2e3%2Fsettings-document-analysis-22-edit-choice.png?alt=media" alt="Edit column window for a Single choice column with five options and their descriptions, Use both and Add option"><figcaption><p>The descriptions are what the AI reads to decide which option fits, so make each one distinct.</p></figcaption></figure>

After the run, these columns get a section of their own on the results page, named after the unit, for example "Documents" or "Outcomes". Each column's block opens on **Values**, which holds **Overview**, **Visualizations**, and **Table**; the counts are units, not documents. A column whose values are all placeholders, such as "Not rated", gets no block.

<figure><img src="https://1178311500-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fi4GBXPMUDxszIpUVOARU%2Fuploads%2Fgit-blob-611ab86d642178295f1cb19c99c4a3fac63d3ba3%2Fsettings-document-analysis-45-column-section.png?alt=media" alt="Participant attributes section with Role open: Values, Overview, a chart counting participants, and the Table of codes"><figcaption><p>Badge 1 is the column and its description, 2 Values, 3 Overview, 4 Visualizations (counting rows), and 5 Table.</p></figcaption></figure>

## A realistic example

> **Scenario:** 3,000 open-ended responses to *"Why did you cancel?"* Select the column, generate a codebook (price, missing feature, switched to competitor, support, bug), and run it. Export the coded CSV, confirm the AI's reasons on the checklist tab, then chart cancellation reasons by [segment](/ailyze-docs/reference/glossary.md#segment) — for example, by plan.

## Next steps

* [Build a codebook](/ailyze-docs/analyze-data/build-a-codebook.md) — tighten the categories the AI uses.
* [Compare groups (segments)](/ailyze-docs/analyze-data/compare-groups-segments.md) — break the counts down by group.
* [Analyze open-ended survey responses](/ailyze-docs/common-research-workflows/analyze-open-ended-survey-responses.md) — the full survey workflow.