Einstein Next Best Action (ENBA) is an out-of-the-box Salesforce Platform feature that provides the possibility to configure business rules and filters that surface the best course of action for any user. This tool can offer an array of recommended actions immediately accessible from within Salesforce.
The goal of ENBA is to display the best action to the right user at the right time — supercharging employees, partners, and customers.
The recommended actions are stored as a standard object in Salesforce called “Recommendations.” They contain text, an Image, any desired custom fields and lookups to select the flows that execute when the recommendation is accepted/rejected by the user.
Using Strategy Builder to funnel recommendations
To surface these recommendations, it is necessary to build a strategy. To do so you would leverage the “Strategy Builder” which is a brand new tool in Salesforce that is identical to the lightning flow builder, just with a different toolset. This new Strategy Builder allows for AI integration (internal and external).
Strategy using Einstein AI
As an example, let’s imagine your business is trying to lower an increasing churn rate. In addition, let’s assume you are already using Einstein Discovery to determine how likely these accounts are to attrit. ENBA can be configured to surface the best course of action when the probability of attrition hits a certain threshold, this way helping the rep to prevent a loss.
Displaying the recommendations to end users
With ENBA, employees have access to these recommended actions by using the “Einstein Next Best Action” standard lightning component on lightning pages. Partners and customers can access these tailored suggestions and increase their productivity using the “Suggested Actions” standard lightning component within a community.
ENBA standard lightning component
You can also surface these same recommendations together with other guided actions in a combined lightning component called “Action & Recommendations.”
Action & Recommendations standard lightning component
Harnessing the power of Einstein AI
ENBA does NOT require any additional licenses. The truth is that AI is not powering the feature in itself and ENBA could be implemented purely as automation with a user interface solution. This means that any level of AI is optional when using ENBA (it is recommended for an enhanced solution).
An example of an implementation where ENBA is able to surface recommendations without the need to leverage AI could be the recommendation of credit card offers. The Sales rep would immediately see the suggested offers which appear based on lead fields, lead status and credit score. While on the phone with the lead they can decide if any aligns with their needs and begin the application process together. In this case, all that is triggering the recommendations to appear would be the data being retrieved from the Salesforce objects. Note that ENBA can analyze cross-object fields before surfacing recommendations (i.e. if a contact’s account has premium support) which allows for a more granular recommendation offering.
Using AI together with ENBA provides actionable items driven by insights that AI extracts from data.
Whether it is using the Einstein Platform or an External source (IBM Watson, Google Cloud Natural Language, etc) Salesforce has an option in “Strategy Builder” to integrate AI with every other piece of the ENBA process.
Strategy builder custom connections (Invocable Apex)
What’s new with ENBA?
ENBA was made generally available in the Spring ‘19 release and at that point it was already being fast tracked due to its potential and its capabilities to enhance the user experience. Here’s a look at the new features and enhancements that were added since then:
- Summer ‘19: ENBA can be used together with other Actions in the Actions & Recommendations component
- Summer ‘19: Available for packaging
- Winter ‘20: Autolaunched flows (background process) are supported
- Winter ‘20: Display recommendations in the Home page
- Spring ‘20: Rejection of the recommendation can also launch a flow
- Spring ‘20: Analyze recommendation strategy metrics (usage/results)
- Spring ‘20: Create Recommendations from the Records of Any (custom and standard) Salesforce Object
