Develop new insights with Einstein Prediction Builder

Predictive insights are the future of business, and within the suite of Salesforce’s Einstein products – there is a tool called Einstein Prediction Builder that all users will want to become familiar with.  Einstein Prediction Builder lets you make predictions about almost any field in Salesforce with just a few clicks. Then you can use the predictions to power a workflow, focus your efforts, and work smarter.  Some of the most common use cases for Einstein Prediction Builder could be centered around Lead Score & whether or not your sales teams are spending time on leads that are likely never to convert.  Or perhaps your Support Team gets a particularly low CSAT score on cases of a certain priority?  Or one of the most common – predicting what customers are likely to churn?  Let’s dive in and start predicting!

EPB1.png

-Pick your object (standard or custom): perhaps you want to focus on which customers are likely to churn?  For instance, this could be the contact or account object where you update if churned status.

 

-Pick your field you want to predict: in the example, if this is a the Churned field – you’ll want to just ensure that there are other records in Salesforce that show some records as churned and some as not

 

-Pick the contributing fields: whichever object you select, you’ll want to then select those supporting fields that help tell the story around a churned customer.  You may select every field or decide to omit certain ones that don’t contribute valuable insights.

 

-Save your selections & Salesforce will create a new field to populate the prediction in:  In the above example, you can create a field such as “Churn score”.

 

-Finally, embed the Einstein prediction component on the page layout of the object you selected.

EPB2.png

Although the setup can be somewhat straightforward – there are some key challenges that need to be considered.  One major challenge is volume of data available to build the prediction around.  You will want to have a large data set for Einstein to review & build its predictions around.  Furthermore, the quality of that data is also key.  If there is a data problem, you will most likely have a prediction problem.  However, if you feel good about the place your data is in & know you have the volume of outcomes that be used for historical perspective – then you are perfectly setup to get started!  As of now, Salesforce does allow each Org the opportunity to build & deploy 1 free prediction….so lets get going!