Next Best Action: Smarter Decisions, Every Time

Why Next Best Action Exists

In Salesforce, users often hesitate when looking at a record: Should I follow up now? Escalate the case? Offer a discount? That hesitation adds up across dozens of users and hundreds of interactions.

Einstein Next Best Action (NBA) eliminates the guesswork. It looks at real-time data, applies your business rules and AI insights, and recommends the most relevant next step—clearly surfaced on screen.

How Einstein Next Best Action Works

1. Recommendation Builder

Create the actions you want to suggest—like “Schedule a Call,” “Offer Loyalty Discount,” or “Escalate Case.” Each recommendation is a record with:

• Label and description.

• Category for grouping.

• Defined action (launch a Flow, update a record, redirect, or trigger Apex).

2. Strategy in Flow Builder

This is the decision engine. Build logic to determine who sees a recommendation, when they see it, and why.

• Use filters, branching logic, and ranking.

• Factor in user roles, record conditions, and AI-driven scores.

• Limit and prioritize recommendations so only the most relevant surface.

3. Flows or Apex Logic

When a user clicks, automation kicks in.

• Use Screen Flows for guided processes.

• Use Apex for advanced logic, third-party calls, or custom actions.

• Leverage the IsRecommendationAccepted flag to branch outcomes depending on whether a user accepts or rejects a recommendation.

Through the Einstein Next Best Action component, users see recommendations directly in Salesforce pages (record, home, app, or Experience Cloud). Accepted or rejected actions are tracked with visual indicators—ensuring clarity and accountability.

How to Set It Up

1. Plan Your Recommendations and Automation
   Decide where recommendations appear, who sees them, and what automation should run upon acceptance.

   
   

2. Build a Flow
   In Flow Builder, design the automation triggered when users accept or reject a recommendation. (Only Screen Flows and Autolaunched Flows are supported.)

   

   
   

3. Create Recommendations

   • Either Manually in the Recommendation object.

   • Or From other data in Flow Builder via the Recommendation Assignment element.

   • Or Automatically with Einstein Recommendation Builder.

     

   
   

4. Create a Strategy in Flow Builder

   • Go to Setup > Flows > New Flow > Recommendation Strategy.

     

   • Use Get Records, filters, and logic (Collection Sort, Decision, Loops).

   • Add recommendations to the outputRecommendations collection for display.

     
     

5. Display Recommendations
   Add the Einstein Next Best Action component to Lightning pages, home pages, app pages, or Experience Cloud sites.

 For deeper technical steps, see the official Salesforce Help Documentation.

Real-World Examples

• Financial Services: Suggest a premium upgrade when a client reaches a balance threshold.

• Healthcare: Recommend preventive care steps during a patient check-in.

• Retail: Display personalized offers based on shopping history.

• Government Services: Help citizens navigate processes with contextual prompts.

Screenshots of Next Best Action in Action 

Final Take

Einstein Next Best Action is more than a suggestion tool—it’s a decision engine. By combining rules, predictions, and automation, it helps teams deliver the right action at the right time. The result? Confident users, smoother workflows, and better customer outcomes.