> For the complete documentation index, see [llms.txt](https://helpcenter.channable.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://helpcenter.channable.com/manage-improve-product-data/categorize-products/ai-product-categorization/how-to-set-up-ai-product-categorization.md).

# How to set up AI product categorization

Set up AI product categorization and review category mappings for your channel.

### <i class="fa-memo-circle-check">:memo-circle-check:</i> Before you start

**You'll need**

* A **Project** with imported items.
* A **Channel that supports** AI categorization.
* Product fields with useful category context, such as `product_category`, title, description, brand, or type.

**Good to know**

* AI categorization is available from the **Core Standard** subscription.
* For channels using Standard AI categorization, rules created through **Generate from field**, **+ Add category rule**, or **Copy from channel** override AI-assigned categories.
* Paused category rules exclude the affected products from categorization.

***

### Set up AI product categorization

Choose the instructions that match your channel and account setup:

* **Improved AI categorization:** From October 13, 2026, new users use this by default for Google, Meta, Amazon, and Bol. It uses one or more category reference fields. See [Improved AI categorization](/manage-improve-product-data/categorize-products/ai-product-categorization.md#improved-ai-categorization).
* **Standard AI categorization:** Use this for other supported channels or when you select Standard AI categorization for Google, Meta, Amazon, or Bol. It uses a single category field. See [Standard AI categorization](/manage-improve-product-data/categorize-products/ai-product-categorization.md#standard-ai-categorization).

{% tabs %}
{% tab title="Improved AI categorization" %}
For Google, Meta, Amazon, and Bol, you can switch between Improved and Standard AI categorization.

<figure><img src="https://501928696-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fu8whzIfip72Kvn3M1bsc%2Fuploads%2FGKtyqZr7VMhXCzu7Gmt9%2FScreenshot%202026-10-13%20at%2008.50.34.png?alt=media&amp;token=04611565-838b-414d-ae96-5661b1aa5232" alt=""><figcaption></figcaption></figure>

{% hint style="info" %}

#### Want to switch to Standard AI categorization?

In **Categories**, turn off **Use improved AI categorization**. The channel then uses Standard AI categorization.
{% endhint %}

{% stepper %}
{% step %}

#### Open the category settings

1. In Channable, go to your project.
2. Go to **Categories**.
   {% endstep %}

{% step %}

#### \[Optional] Copy categorization from another channel

Use existing mappings from another channel as additional context.

1. In **Copy from channel**, select a channel with suitable existing mappings.

{% hint style="info" %}

#### Using copied mappings

Channable uses the selected channel's existing category mappings as additional context during AI categorization.
{% endhint %}
{% endstep %}

{% step %}

#### Choose category reference fields

Select the product fields that AI uses to identify the right category.

1. In **Category reference field(s)**, select the field that best describes each product.
2. Click **Add another field** to add context, such as title, description, or brand.

Select one or more fields to help match your items to categories in your channel. `category` is selected by default.

{% hint style="info" %}

#### No category field?

Descriptive fields, such as title or description, can give Channable enough context to assign categories.
{% endhint %}
{% endstep %}

{% step %}

#### Start categorization

Start AI categorization and monitor the results as processing progresses.

1. Click **Start categorization**.
2. Wait until the category status shows that processing is complete.

{% hint style="info" %}

#### Monitor categorization progress

The **Categorized items** and **Uncategorized items** tabs show the number and percentage of items in each group. An upward arrow identifies categorized items. A downward arrow identifies uncategorized items.

The counts and percentages update as Channable categorizes your items.
{% endhint %}

{% hint style="success" %}

#### Categorization result

When categorization completes, you see one of these messages:

* **Minimum number of items categorized:** You have categorized enough items to continue. Go to the **Rules** step.
* **All items categorized:** You have categorized all items. Go to the **Rules** step.

For Amazon only:

* **All items have a product type:** Go to the **Rules** step. Set up browse nodes later.
  {% endhint %}
  {% endstep %}

{% step %}

#### Review and correct categories

Review category paths and correct items that need a different mapping.

1. In **Categorized items**, click the expand arrow next to a category path to view its assigned items. Each category path shows the number of assigned items, such as `75 items`.
2. Select individual items, or select all items, using the checkboxes.
3. Click **Change category**.
4. In **Change to category**, review the ID filter. It follows the format: **When** `[ID]` **is equal to** `[value]`, **add item to**.
5. Search for and select the correct category. You can search in another supported language. For example, search `spijkerbroeken` to find jeans.
6. Click **Save and create rule**.

Channable creates the category rule and adds it to the **Category rules** tab. The tab shows the number of configured rules.

{% hint style="info" %}

#### Assign Amazon browse nodes

For Amazon, the bottom of **Categorized items** lists category paths, browse-node dropdowns, and affected item counts. Select a browse node from the dropdown for each category path.

Browse nodes do not block you from continuing your setup. You must assign them before you activate your Amazon channel.
{% endhint %}
{% endstep %}

{% step %}

#### Categorize uncategorized items

Create category rules directly from the **Uncategorized items** tab.

1. Open **Uncategorized items**.
2. In the rule builder, select the project field that identifies the items.
3. Select a statement, then enter the value to match.
4. Click **Categorize items**.
5. Search for and select the category to assign.
6. Click **Save**.

For example, create this rule: **If** `description` **contains** `t-shirt`, set the category to `Clothing & Accessories > Clothing > T-Shirts`.
{% endstep %}

{% step %}

#### Rerun AI categorization

Rerun AI categorization after you add or change category reference fields.

* Click **Rerun**.

{% hint style="info" %}

#### Add more context for low coverage

If AI categorization covers fewer than 50% of your items, Channable prompts you to add more input fields. Add descriptive fields, such as title, description, brand, or product type, then click **Rerun**.
{% endhint %}
{% endstep %}
{% endstepper %}

Click **Rules** to continue your channel setup. Complete the **Build** step after rules.
{% endtab %}

{% tab title="Standard AI categorization" %}
Use these instructions for channels that use Standard AI categorization.

<figure><img src="https://501928696-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fu8whzIfip72Kvn3M1bsc%2Fuploads%2FFZKqTDtJAjYM1FMJayuf%2FScreenshot%202026-10-13%20at%2008.51.18.png?alt=media&amp;token=f70d3e50-d1b7-4c4b-82a8-52e033a48b45" alt=""><figcaption></figcaption></figure>

{% hint style="info" %}

#### Want to switch to Improved AI categorization?

For Google, Meta, Amazon, and Bol, turn on **Use improved AI categorization** in **Categories**. The channel then uses Improved AI categorization.
{% endhint %}

{% stepper %}
{% step %}

#### Set the category field

Choose the product field that AI uses as the basis for category suggestions.

1. In Channable, open your **Project**.
2. Click **Channels** > **+ Create channel**, or select an existing channel.
3. In **Categories**, click **Set category field**.
4. In **Set category field**, select the field that AI categorization uses, such as `category`, `title`, or `description`.
5. Click **Save**.

{% hint style="info" %}

#### Changing the category field

The `category` field applies to every channel in the project that uses Standard AI categorization. Select one field that works for all those channels.

The `category` field is selected by default. Changing the field reruns categorization and resets automatically generated categories for those channels.
{% endhint %}
{% endstep %}

{% step %}

#### Review the suggested categories

Check the suggested mappings and identify items that need manual categorization.

1. In **AI product categorization**, review the category mappings.
2. In **Mapped from category field**, check the values from the field you selected. For example, if you select `title`, a value might be `American Outfitters girls dress Sunshine`.
3. In **To category**, check the assigned channel category. In this example, the items might map to `Clothing & Accessories > Clothing > Dresses`.
4. In **Count**, check how many items use the mapping. In this example, `6` items are mapped. Click the number to inspect them.

**Uncategorized items** lists products without a category. Inspect them there, then use **Manual categorization** to create category rules.
{% endstep %}

{% step %}

#### Choose a manual categorization method

Select the method that you want to use to map uncategorized items (or re-map AI categorized items).

1. Click the **Manual categorization** tab.
2. Choose one of the following methods:

{% content-ref url="/spaces/u8whzIfip72Kvn3M1bsc/pages/92d971c4c1fe33c75fa11a9b94dd5b3b129ac6f0" %}
[Generate category rules from a field](/manage-improve-product-data/categorize-products/manual-categorization/generate-category-rules-from-field.md)
{% endcontent-ref %}

{% content-ref url="/spaces/u8whzIfip72Kvn3M1bsc/pages/909f27a16b099b34ceeeeb3d7e3b5671e0dacfb9" %}
[Manual categorization](/manage-improve-product-data/categorize-products/manual-categorization.md)
{% endcontent-ref %}

{% content-ref url="/spaces/u8whzIfip72Kvn3M1bsc/pages/96d30a28ee1e062c116b69bf05ae942702d75675" %}
[Copy categories from another channel](/manage-improve-product-data/categorize-products/manual-categorization/copy-categories-from-another-channel.md)
{% endcontent-ref %}
{% endstep %}
{% endstepper %}

{% hint style="info" %}

#### Category rules override AI categorization

Rules created through **Generate from field**, **+ Add category rule**, or **Copy from channel** override AI-assigned categories.
{% endhint %}

### Start again

Remove the category field to clear auto-categorized products. The categorization overview then shows that no category field is set.
{% endtab %}
{% endtabs %}


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://helpcenter.channable.com/manage-improve-product-data/categorize-products/ai-product-categorization/how-to-set-up-ai-product-categorization.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.