Recommendations help deliver personalized content to users, enhancing engagement and driving conversions. These recommendations are powered by data-driven insights that align with user behavior and preferences.
The Import Recommendations page allows you to efficiently manage all your recommendation data imports. Blueshift supports file uploads and data warehouse imports, providing a streamlined process to ensure your campaigns deliver personalized, data-driven content. This guide will help you navigate the import process and manage your recommendation feed imports effectively.
Go to recommendation feeds
- Navigate to Customer AI from the left-hand navigation panel and click on Recommendations.
- Click the Feeds tab at the top of the screen to access the imported recommendations index page.
You will find the following features on the screen:
- Feeds index - Quickly view all your existing content feeds with details like the feed name, type (user to product or product to product), processed records, and timestamps for when they were last updated.
- Search and filter - Use the search bar to locate specific feeds by name or apply filters to narrow the list based on authors.
- Feed details - Click on any feed to view in-depth information about its import status, processed records, and error reports.
- Create a new feed: Click the + Content Feed button in the top-right corner to start a new import.
Supported feed types
You can upload custom recommendations directly into Blueshift as recommendation feeds. The recommendations must be in the form of a CSV file. The following types of feeds are supported:
- User-to-product: Recommendations for each user can be imported using one of the user identifiers and one or more product IDs from the catalog. For example, this is useful for personalized promotions.
- Product-to-product: Recommendations that suggest related products based on user interactions. For example, people who view or buy a product might also consider other related products.
Data format
After you prepare a CSV file containing the recommendations data, you can import it to Blueshift by creating an import feed.
- Include either email or customer_id as the customer identifier for user-to-product recommendations.
- Separate multiple products with a delimiter (e.g., a comma or pipe symbol).
- Click the Download sample CSV link in an import task to review the required format.
Sample data for user-to-product recommendations
Product IDs | |
---|---|
testuser1@domain.com | ITEM_0000 | ITEM_1111 |
testuser2@domain.com | ITEM_222, ITEM_333, ITEM_5555 |
Sample data for product-to-product recommendations
Source Product ID | Product IDs |
---|---|
ITEM_123 | ITEM_234, ITEM_345 |
ITEM_234 | ITEM_123 | ITEM_345 | ITEM_456 |
Choosing your data source for import
When you create a new content feed, you'll be prompted to select the data source:
-
File imports:
- Direct file upload (CSV format)
- Upload via S3 bucket
- Upload via SFTP
-
Data warehouse imports:
- Snowflake
- Databricks
- Google BigQuery
Next steps
After selecting your data source, you'll be guided through configuration steps tailored to your import type. Whether mapping fields from a CSV file or setting up a data pull from your warehouse, the intuitive interface ensures a smooth workflow.
For detailed instructions on each import method, refer to:
Comments
0 comments