Einstein Recommendations – How it Works?

Einstein Recommendations – How it Works?

Sharing is caring!

Marketing gains superpowers with Einstein for Marketing Cloud, with extra advantages. Einstein Engagement Scoring gives a score to every single subscriber based on the level of engagement with your brand. This data is used in personas and the personas are used for targeting the audiences. The customers who are loyal to your brands are then sent limited offers or those with lesser engagement are all put into special engagement programs.

Einstein segmentation uses AI for segmentation purpose. Einstein Vision for Social uses AI for looking at images or social posts for product discovery and reaches out to the consumers. Einstein Journey Insights tell you about what the customers go through their journey.

In this post, we will throw some light on Einstein Recommendations.

Einstein Recommendations and Types of Data

This speaks about not just product recommendations but also about content recommendations. It captures the behavioural data on consumers and using them for unified profiles. Moreover, AI is used for building personas, finding out what their affinities are. Data for all the customers are gathered to predict the preferences of individual customers. You can then communicate with your customers with easy-to-use drag and drop tools. Einstein Recommendation gives you the following data:

 

  • Behavioural Data

  • Affinity Data

  • Recommendation Engine

 

Behavioural Data


It is about all behavioural data for the consumers.

 
Affinity Data

The preferences of the consumers, the topmost ones.

 Recommendation Engine

Based on the preferences, the recommendations are made.The recommendation logic is packaged into displays across channels. The Business rule engine customizes the recommendation types presented, based on context.

How to Use this data?

The data from Einstein Recommendations is used in the following:

 

Content Delivery

It is about automating the content for various channels – web, mobile and email.

Segmentation
The behavioral and the affinity data is also available inside the Marketing Cloud. This helps in segmentation. This way you can create marketing campaigns, based on this data and use it for marketing differently with high value clients, for higher-end products.

Definition
Based on the behaviour data, events are brought inside Marketing Cloud System for consumers losing interest about your products for placing them inside reengagement journeys.

Now, we will find out how Einstein Recommendations is used by the shoppers.

How To use Einstein Recommendations?

To get started, you need to have a product catalog, or we can even import a catalog. We also have a streaming update features for embedding the catalog items or status for items, So, the products not in stock can be conveyed to the customers and these are then not recommended in future. Once the catalog is in place, there is a tracking code. In order to use Einstein Recommendations in Marketing Cloud, we need to do the following:

 

Configuration

 

We use the personalized builder for configuration that sets the choice for type of recommendations – Product or Catalogs. Here we specify the required product fields or the customized fields. The localization is ensured by the localization support option. While tracking users, we can specify the events to track too. It offers a summary of all the choices made.

How to use the Catalog?

This sets the option on How to use the Catalog? Whether to use streaming updates or use batch updates. If the catalog is ready, we can even upload it here. We specify the tags to be used for tracking the detailed view or the category view.

Recommendations Scenarios

This is a feature in the Personalization Builder. We configure the scenarios say “bought bought” where a product is bought and the client is interested in some other products too.

 

Waterfall Predictions

We have the option to enable Waterfall Predictions in the Personalized Builder. on the number of recommendations and the order to be followed. It states the top priority scenario as well as those with lowest scenario. The layout to be presented to the client is also selected here. You need to set a maximum number of recommendations and the system returns the matching items until the threshold value is reached.

 Rule Manager

 

You can refine Einstein Recommendation for web or email by using the Rules Manager of Personalization Builder.The Rules in the Rules Manager is to funnel down the preferred recommendations. The rules are created for all web and email recommendations. You can simply configure a rule for a better output from Einstein Recommendations. There is a way to configure advanced rule settings. There is a way to validate rule configuration by review of the rule text, consistent with your expected result, before saving the rules.

Email Recommendations

This is about recommendations for clients, for sending emails.  We can edit the content to be sent here. We have the option to drag and drop the Recommendation block.  We can pick on the look and feel of the display and the scenarios here.The layout size can be selected from this section. There is an option to preview the document too.

Summary

Einstein Recommendation gives you the data on Behavioral Data, Affinity Data and Recommendation Engine. The behavioral data is about  the data on the consumer behavior and the topmost preferences are given by the affinity data. The above types of data are available within Marketing Cloud. This helps in the segmentation of clients.

 

You can use Einstein Recommendations to state “How to use the Catalog?”. Here we state whether to use streaming updates or use batch updates of catalogs. The various scenarios and the order of recommendations. You can also edit the recommendations for the email. The same layout can be previewed too.

Ajay Dubedi

Ajay Dubedi

CEO | Founder
"Ajay Dubedi, the founder and CEO of Cloud Analogy, is a prominent Salesforce Sales, Service, and Marketing cloud Consultant with a rich expertise in handling challenging business models. Ajay has assisted and implemented solutions in industries comprising Banking, Health Care, Networking, Education, Telecommunication and Manufacturing. For the last many years, Ajay has been instrumental in passing on his vast knowledge among his colleagues and strongly believes in harnessing an atmosphere of encouragement, empowerment, and mutual advancement. Ajay is globally acclaimed for his extensive experience in APEX Programming, VisualForce pages, Triggers, Workflows, Page Layouts, Roles, Profiles, Reports & Dashboards."

Leave a Reply

Close Menu