> ## Documentation Index
> Fetch the complete documentation index at: https://moengage.com/docs/llms.txt
> Use this file to discover all available pages before exploring further.
# Trending Items
> Create recommendations that surface trending items within a specific catalog attribute, such as category or location, using aggregated user interaction data.
# Overview
Trending Items is a Merlin AI-powered recommendation model that identifies items gaining traction based on aggregated user interactions. MoEngage groups interaction events such as purchases, views, and wishlists over a defined time period to surface what is trending within a catalog attribute you choose, such as category, subcategory, or city. You can then link those trending results to individual users based on their recent activity or profile attributes.
**Prerequisites**
Before you create a Trending Items recommendation, make sure you meet the following requirements:
* You've enabled Advanced Recommendations for your account.
* You've integrated user action events with MoEngage and ensured each event includes a unique item ID.
* You track and map the relevant user-item interaction events to standard MoEngage events.
* Your account contains an active catalog.
* For Attribute Trend, your catalog includes at least one categorical item attribute, such as `category` or `city`.
For more information, refer to [Map User Actions](/user-guide/content/recommendations/prerequisites/map-user-actions-settings) and [Catalogs](/user-guide/content/recommendations/prerequisites/catalogs).
## Use Cases
Trending Items supports the following use cases:
* **Category-specific trends**: When a user browses the Electronics section, show items trending in Electronics. When they browse Fashion, show trending items in Fashion.
* **Regional trends**: Surface items trending in a user's city or region based on a location attribute in your catalog.
* **Fallback recommendations**: Use Trending Items as a fallback when Frequently Viewed Together or Frequently Bought Together recommendations have insufficient interaction data.
## Trend Types
Trending Items supports two trend types: **Global Trend** and **Attribute Trend**.
1. **Global Trend**: Aggregates interaction data across all items in your catalog to surface items trending at the catalog level, regardless of category or attribute. Configurations include:
* **Generate trends, ranked by**: The interaction metric used to rank items such as Best sellers, Most viewed, Most wish-listed, Most cart additions, or Top revenue generators.
* **In the last**: The period over which interactions are aggregated, such as 24 hours, 7 days, 14 days, or 30 days.
2. **Attribute Trend**: Identifies items trending within the value of a specific catalog attribute. Instead of surfacing trending items globally, it shows what is trending within the attribute value relevant to each user, such as the category they last browsed or the city stored in their profile.
## Trend Configuration
You can use the following options to configure an attribute trend:
| Field | Description |
| -------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------- |
| **Generate \[attribute] trends** | The catalog attribute to group trends by, such as `category`, `subcategory`, or `city`. |
| **Ranked by** | The interaction metric used to rank items: Best sellers, Most viewed, Most wish-listed, Most cart additions, or Top revenue generators. |
| **In the last** | The period over which interactions are aggregated: 24 hours, 7 days, 14 days, or 30 days. |
# Trend Source
After you configure the trend, MoEngage determines the relevant attribute value for each user based on their profile. Select one of the following options to define the attribute value:
* **Custom value**: Enter a fixed attribute value. MoEngage shows items trending in that specific value for all users targeted by this recommendation. Example: If your catalog attribute is `Category` and you set the custom value to `Electronics`, **all** users will receive recommendations for trending electronics items, regardless of their individual locations or browsing history.
* **User property**: Map to a dynamic user property that stores the relevant attribute value for each customer. MoEngage uses each individual user's real-time property value to personalize their results.\
Example: If your catalog attribute is `City` and you map it to the user property `Home City`:
* User A (with `Home City` set to "Chicago") will see items trending in **Chicago**.
* User B (with `Home City` set to "London") will see items trending in **London**.
# Create a Trending Items Recommendation
Admin, Manager, and Marketer roles can create recommendations.
To create a Trending Items recommendation, follow these steps:
1. On the sidebar menu in MoEngage, hover over the **Content** menu icon 
. The Content menu is displayed.
2. Click **Recommendations**. 
The Recommendations page is displayed.
3. Click **+ Create recommendation**. 
The Create recommendation page is displayed.
4. Under **Predictive**, select **Trending Items** and click **Next**. 
5. In the **Recommendation name** box, enter a name for the recommendation.
6. In the **Recommendation description** box, enter a description (optional).
MoEngage pre-populates the **Catalog** box with the catalog selected for all AI recommendations in your account. The **Catalog** field is disabled and cannot be changed after selection. To use a different catalog, contact your customer success manager (CSM).
7. Under **Trend configuration**, select a trend type:
* **Global trend**: Aggregates interactions across all items in your catalog.
1. Click the **Global trend** card.
2. In the **Generate trends, ranked by** list, select the interaction metric that MoEngage uses to rank the trending items (for example, Best sellers).
3. In the **In the last** list, select the time period over which MoEngage aggregates interactions (for example, 7 days). 
* **Attribute trend**: Surfaces items trending within the value of a catalog attribute.
1. Click the **Attribute trend** card.
2. In the **Generate \[attribute] trends** list, select the catalog attribute to group trends by, such as `category`.
3. In the **Ranked by** list, select the interaction metric.
4. In the **In the last** list, select the aggregation period.
5. Under **Show trends to user, where**, select the catalog attribute and the operator. Choose **Custom Value** to enter a fixed value, or **User property** to map to a user profile property. Optionally, select **Case Sensitive** to match the value exactly. 
8. Optionally, turn on **Filter items by user actions** to control which items appear in the results:
* **Select user action to keep items**: Include only items that appear in both the recommendation results and the specified user action criteria. Click **+ Filter** to add the action, time operator, and value.
* **Select user action to remove items**: Exclude items from the recommendation results based on the specified user action criteria. Click **+ Filter** to add the action, time operator, and value. 
9. Optionally, turn on **Filter items by item attributes** to filter the results based on catalog data. Under **Item where**, select the product attribute and the operator, and then choose **Custom value** or **User property** to define the value. To match the value exactly, select **Case sensitive**. Click **+ Filter** to add the condition. 
10. Click **Create**.
In the **Test recommendation** section, **Show results** stays disabled until MoEngage finishes processing the recommendation. A banner is displayed: "You will be able to test recommendations after processing is complete. Processing will complete in 2 hours."
# Test the Recommendation
Trending Items results can differ for each user based on their recent activity, so MoEngage doesn't auto-populate them. After you create the recommendation, MoEngage takes about 2 hours to process it. After the processing is complete, follow these steps to preview the output:
1. Go to **Content** > **Recommendations**.
2. Find your recommendation in the list and click the edit icon under **Actions**. 
3. In the **Test recommendation** section, type a MoEngage user ID in the **Show recommendations for** box.
4. Click **Show results** to preview the recommended items for that user. 