You are here:
Einstein Content Selection
Einstein Content Selection is a no-code, open-time experimentation tool for personalized content. Because Einstein selects content at open time, it continues to evaluate for winning content while also experimenting to ensure that click-through rates are optimized.
Before starting, here are a few key concepts to note.
- Asset catalog—Use the set of images in a catalog in your experiments.
- Asset class—To determine how the images are used, add assets to an asset class. Asset classes are frequently used to determine email template placement, such as the banner, main call to action, or the footer.
- Asset attribute—Categorize your images, personalize your content, and analyze click-through rates using attribute values. For example, an asset attribute called Loyalty_Level has three values: Silver, Gold, and Platinum. When you pair the attribute with consumer profiles, Einstein can match the correct set of experimental images using the customer’s loyalty level.
- Selection rule—Set up business rules to align Einstein with your business goals and intentions, such as how frequently content can be reused.
Use Case Examples
A marketer for an apparel company uploads a group of product image assets each week. Each product image asset features an item that’s on sale that week. The marketer creates an asset class called WeeklySalesAssets and assigns the product image assets to it. The marketer uses StartDate and EndDate to make sure that the latest weekly sale item is shown at open time.
A national grocery store chain decides to use email marketing to distribute weekly coupons. The marketers create an asset class called CouponFeature. Each week, marketers upload image assets for the coupon products. Each image includes the CouponFeature asset class. The marketing team uses StartDate and EndDate to ensure that the assets are available for selection in emails only during the week that they’re on sale. The marketers also check Einstein Content Selection analytics to find the week’s highest performing assets. They’ve enabled Einstein Content Tagging to monitor top-performing sale types and discovered that coupons for cleaning products outperformed others. Then they compare the results with sales figures to understand the relationship between the items that customers click and the items that they buy.
Getting Started
Configure a few settings, and supply Einstein with content assets. Import bulk image assets using one of these methods.
- Import an asset catalog file that contains metadata and image links that you use in marketing content.
- Use the Einstein Content Selection API.
- Import images from Content Builder or directly from your desktop.
Begin sending by adding an Einstein Content block to an email. Einstein then optimizes asset selection for the highest click-through rate at open time. If you decide to add customer profile information, each email is further customized for the subscriber using the assets that you provide. You can also add content selection and fatigue business rules to help Einstein align with your business goals.
An Einstein Content block can overwrite your edits. To customize, you can drag an Einstein Conten block onto a temporary email, choose the asset class, and switch to code view. Copy the code, then paste it into a new content block's code view. You can then modify the new block to customize its look and feel. But be cautious: Copying and pasting HTML can lead to errors like bad links, and reusing the same HTML block can cause issues due to duplicated block IDs. The ECS block automatically assigns unique IDs, minimizing these risks. If you still choose to use a custom HTML block, carefully review your code to avoid errors.
As subscribers engage with your messages, Einstein analyzes click results to find out which assets spark interest in customers. Using real-time analysis, Einstein finds the assets that prompt customers to click a link in the email. As analysis proves that an asset continually contributes to increased click rates, Einstein selects that asset more frequently when customers open their email.
Related Content
- Set Up Einstein Content Selection
Enable the Einstein Content Selection message block so that you can add it to your messages in Content Builder. - Create and Upload an Asset Catalog File
Give Einstein the images that can be selected as content assets in your messages. Provide an asset catalog .csv file that contains a URL and metadata for each content asset image file that Einstein can select. We provide a template to guide you. You can use Einstein Content Selection in your email messages after uploading assets. - Manage Einstein Content Selection Assets
Manage the asset pool you created on the Asset Pool page after you add assets to Einstein Content Selection. You can add assets to the pool in several ways. Upload an asset catalog file, upload a file from your desktop, import the asset from Content Builder, or add the asset through an API The assets available to you are the image assets that Einstein Content Selection chooses from based on the asset class of the content block. If you’re importing profile information, the selection is based on how the image attributes align with the subscriber profile attributes. - Manage Asset Classes
You can add, modify, archive, and restore asset classes. - Sync Your Consumer Profile Data Extension
If your organization has consumer or customer profile data that you want Einstein Content Selection to use when making selections, sync the sendable data extension that contains your customer profile data. When the data extension gets updated, sync it again so Einstein has the latest profile data. Or set up automatic daily profile syncing so you don’t have to. - Set Up Content Selection Rules
If you want Einstein Content Selection to observe rules about what content to select, set content selection rules. A content selection rule governs when some assets can and can’t be selected and included in a message. - Set Fallback Assets
Choose an image asset that Einstein Content Selection displays when a matching asset can’t be found. Pick an image for each asset class in your asset catalog file. Even if your fallback assets are unavailable, Einstein Content Selection selects an image if possible. - Change or Remove a Fallback Asset
If Einstein Content Selection can’t find an asset that matches your content selection rules, it displays a fallback image. You can change or remove the fallback image for each asset class. - Find Out Why an Asset Was Selected
Search by subscriber key to review the details of an asset that was selected because of its likelihood to engage a specific subscriber. Find out why Einstein Content Selection didn’t select other assets. - Einstein Content Selection Performance Analytics
Discover which aspects of your content images resonate with your audience. View performance over time by asset class, by asset attributes including Einstein’s tag and tone attributes, or by job. You can also get details about assets that Einstein doesn’t select. - Einstein Content Selection Asset Distribution
Import more assets and keep an eye on your asset inventory’s status. Find an overview of assets and their distribution. Monitor how many assets are active or expired, and how many assets expire soon. - Daily Selection and Click Events Report
The daily events report is a summary of subscriber content selection and click events. Use the report in Einstein Content Selection to track subscriber views and engagement. Reports are retained for 90 days. - Use Local Weather Conditions to Target Assets
You can target assets based on the weather forecast and conditions for a subscriber’s location. For example, you can promote sunscreen when the Ultraviolet (UV) index is high or promote raincoats when there’s a 75% chance of rain in the next five days. - Einstein Content Selection and Testing Model Card
The model in this card analyzes message content performance to optimize email engagement.



