ArticlesBite Size Analytics: Best Practices for Using Next Best Actions

For any feedback or requests for future Bite Size Analytics, post in the Reporting Chatter group.

Next Best Action Basics

There are 4 main aspects to Next Best Action:

• Strategy: The strategy you define looks at all of the relevant recommendations available and filters them out to find and prioritize them based on criteria. You can also define how often recommendations appear if they are rejected or accepted.
• Recommendations: Recommendations are records that define the flow that should be launched when accepted. They let you define images, titles, descriptions and button attributes. They can be defined. They can also be dynamic.
• Embedding: Recommendations can be surfaced on records and app pages. When you embed you define which parts of a recommendation are displayed in terms of images for example.
• Flows: Flows launch processes. These can be screen-based to take you through a series of steps such as inserting or updating a record. They can also be used to launch a process such as sending an email.

Enabling the Functionality

Einstein Next Best Action and Flow are enabled automatically in every environment. This means that the below scenario can be setup without any upgrades to packages or requests to Salesforce to enable features.

Recommending a Change Request if close to Budget

The example that this byte will guide you through creating is to have a next best action recommend you create a budget change request when billings are close to bookings. This can be placed anywhere on the project record and the criteria can be customized to your environment.

Step 1: Creating a Flow

The first place to start with a NBA is the flow for the process you wish to recommend. In the example we'll be working through this is the creation of a record. Follow the video below on creating a flow.

Step 2: Creating a Recommendation

After creating the flow above you should create a Recommendation record. These can be created from the recommendation tab and should include the flow you created in Step 1 as the action. Only flows are supported through NBA. An example recommendation based on the above flow would be:

In the above, we have not included an image but you can provide one if desired (1000px x 380px recommended). Custom fields can also be added to this object if you want to give weighting and expand the strategies later on.

Step 3: Creating a Strategy

After creating a flow and recommendation, follow the below video on creating a strategy. This defines the recommendations that will appear and the criteria for that to happen. This process can be customized and many recommendations can be loaded for an object. When loading multiple recommendations it can be useful to add custom fields to the recommendation object in order to sort by weighting as only 4 recommendations can be presented to users at once.

Step 4: Embedding your Next Best Action

Now that you've created your flow, recommendation and strategy, the final step is to embed the lightning component in your page layout. It's position is entirely up to you but you must have administrator permissions.

And there you have it, an Einstein Next Best Action in your environment, giving users recommendations based on the context of the record being viewed. The example above can even be repeated on the opportunity record also.

Certinia are always exploring the latest technologies, from Tableau CRM (formerly Einstein Analytics), Einstein Discovery, and now Einstein Next Best Action. The Next Best Action technology allows criteria-driven recommendations to be displayed to users so that actions can be taken. There are some great trailheads that go through examples of Next Best Action on trailhead that are worth exploring.