Start tracking your progress
Trailhead Home
Trailhead Home

Automate with Einstein Content Tagging and Selection

Learning Objectives

After completing this unit, you’ll be able to:

  • Activate and use Einstein Content Tagging.
  • Create an Einstein Content Selection content block.

Personalized Content

As a marketer, you know that customers want personalized email. Easier said than done, right? In this unit, you meet some Einstein features that can help you quickly enhance your dynamic content and provide customers with the customized experience they crave. Win-win! 

Easy and Automatic Content Tagging

You’ve spent time creating great content, you’ve uploaded images of all your products, and purchased countless stock photos—but you can’t find the one you need when you need it. It’s time to stop wading through your images and activate Einstein Content Tagging.

Einstein Content Tagging automatically applies searchable tags to image files (jpg or png) in your Marketing Cloud account. These tags are different than manual tags that you can add to files. To differentiate between the two, a prefix (that you choose) is added to each Einstein tag to distinguish them from tags added manually. 

Why use intelligent content tagging? To save time, be more productive, and frankly—why not? You can even fine-tune this feature to better serve your needs.

Note

Note

Einstein analyzes and tags new images once every 24 hours. How’s that for smart?

Let’s get tagging. 

  1. In the Marketing Cloud Setup app, expand Einstein and click Einstein Content Tagging.
  2. Click Configure.
  3. Enter the maximum number of tags per image (up to 25).
  4. Enter a Prefix for the tag.
  5. Select the Confidence Level between 0.5 (lowest) and 1.0 (highest).
  6. Click Save.
Note

Note

Use a higher confidence level to improve the accuracy of the tags applied to your images. Use a lower confidence level to broaden the scope of tags included.

How’s it work? It’s Google magic. This feature is powered by Google Vision, which analyzes the pixels and patterns of your image and searches its extensive repository for matching patterns in previously tagged images. Vision’s confidence level is determined by how often it has analyzed similar images and the distinctiveness of the tags applied. So more images = better tags. Now that your content is tagged and easy to find, you can focus on choosing the best content for the job. Einstein can help with that, too.

Take the Guesswork Out of Choosing Content

Picture this: You’re hungry and you’re meeting a friend for dinner. You ask your friend, “Where do you want to eat?” And your friend says, “I don’t know, where do you want to go?” And you reply, “I don’t know. What sounds good? Sushi? Pizza? Italian?” And you go back and forth until you can barely hear each other over the rumbling of your stomach. Sound familiar? Wouldn’t it be nice to just press a button and let someone else decide? Better yet, what if that decision was made based on your location and the types of restaurants you’ve visited before? 

Well, taking the guesswork out of decision making is exactly what Einstein Content Selection does. It may not help you with dinner decisions (sorry to get your hopes up), but it can help you with content decisions. It works by selecting the most engaging content using a mix of rules with neither code nor AMPscript (Marketing Cloud’s proprietary scripting language) needed. Through machine learning and user behavior (like how many clicks content has received from early openers), the application selects the perfect asset from your content library at the time of open. 

Why is this a game-changer? Because AI can help improve your click to open rates in real time, based on what is working with your unique subscribers. Which is essentially real-time content testing, without any intervention from you. 

Get Started with Einstein Content Selection

You had me at real-time. Here’s how you can set up Einstein Content Selection. First, navigate to Setup in Marketing Cloud. Click Einstein and then Einstein Content Selection. Enable this feature to use the Einstein Content Selection block within Content Builder. 

Next, you need to create an asset catalog file that contains metadata about (and links to) the images you use in marketing content. If you decide to add customer profile information, you can help Einstein make more educated choices. Plus, you can add content selection and fatigue rules (these prevent a customer from seeing the same image over and over) to help guide Einstein. 

  1. Create an asset catalog that includes images and metadata that can be used in content selection.
  2. Under Personalization Builder, select Einstein Content Selection.
  3. Click Upload File to upload your asset catalog.
Note

Note

Einstein Recommendations uses a separate catalog to drive the recommendations engine. Learn more about the Einstein Content Selection catalog requirements on the help page: Create and Upload Asset Catalog.

If you would like to link an existing data extension (that has customer preference data) to content selection, click Sync Data Extension. Any time the synced data extension is updated, it needs to be updated in Personalization Builder as well.

You can also identify rules for how often content will or won’t show to the same customer under Content Selection Rules.

Once you have this configured, follow these steps in Content Builder.

  1. Drag the Einstein Content Builder SDK into your email.
  2. Select the asset class identified in your catalog.
  3. Once completed, click Done Editing.

Einstein Content Selection content block with running shoes.

Note

Note

Be sure to set a fallback asset (default content) if no data has been collected.

Once configured and you have sent your email, you can review and analyze your top-performing content and make adjustments based on that data. 

Einstein Content Selection asset performance analytics.

Auto tags? Check. Auto content? Check. With a little help from Einstein, you just entered the world of intelligent personalized content. Now, you have what you need to take personalization even further with Einstein Recommendations. 

Rights of ALBERT EINSTEIN are used with permission of The Hebrew University of Jerusalem. Represented exclusively by Greenlight.

Resources