Get Started with Einstein Recommendation Builder


#AwesomeAdmins, if you know about Flow, you know the power of automation. In fact, ever since Salesforce launched in 1999, our focus has been on automation — beginning with transforming sales.

Over the years, we’ve built low-code tools to help you automate business processes and actions, such as Flow and Einstein Next Best Action. Now, we’re empowering admins to bring AI into their workflows with a new capability: Einstein Recommendation Builder.

Introducing Einstein Recommendation Builder

Einstein Recommendation Builder is the industry’s most easy-to-use recommendation engine. As Salesforce Admins, you can use Einstein Recommendation Builder to build personalized recommendations in context with clicks, not code.

Watch the Trailhead LIVE episode below for an overview including use cases and a live demo, and keep reading for an overview of recommendations, best practices, and more.

What are recommendations?

Recommendations are fairly common in the consumer space. You may have seen these while shopping with your favorite online retailer or browsing for movies or music. For example, if you’re a Netflix customer, Netflix will share recommendations with you based off of factors such as your viewing history, how you’ve rated titles, and other customers’ preferences.

Einstein Recommendation Builder brings a similar recommendation engine capability into the Salesforce Platform that can be leveraged for CRM applications across any domain and vertical.

Here are some of the recommendations admins can build with Einstein Recommendation Builder.

Use Case Example
Upsell/Cross-sell Recommendation Recommend top products that are best fit for a Contact/Account.
Next Best Offer Recommend best offer/promotion to new or existing customers.
Next Best Action Recommend best solution/action based on customer issue.
Field Service Work Order Enrichment Recommend relevant parts to carry for service visit by field technician.
Candidate Recommendation Recommend relevant candidates who are most likely to accept the job, or vice versa.
Customer-Partner Matching Recommend most relevant partner solutions to your customers.
Campaign Recommendations Recommend campaigns to audiences to improve email response rates.
Jobs Recommendations Recommend relevant candidates who are most likely to accept a job to improve placement rates.
Case Resolution Recommend best solution to cases to improve CSAT & time to resolution.

What does this mean for Salesforce Admins?

Admins can help their users solve business problems fast by deploying personalized recommendations in real time. Using clicks, admins can create intelligent recommendations quickly and leverage the power of low-code tools, like Einstein Next Best Action and Salesforce Flow, to automate those recommendations.

Einstein Recommendation Builder makes it easy for you to build AI-powered recommendations, allows you to seamlessly combine machine learning with business rules, and provides model transparency with metrics.

How to build and deploy recommendations

Let’s walk through a scenario together.

Here, you’re looking at a new service work order from a customer.

New service work order in Salesforce

But the challenge is that this work order doesn’t have the right product parts to resolve the customer’s issue. Because of that, we’ll use Einstein Recommendation Builder to recommend the product parts most likely needed for the work order.

Here are the steps:

  1. Go to Setup in your Salesforce org, and search for Einstein Recommendation Builder.
  2. Select the Salesforce object that contains what you want to recommend. In this case, it’s the Product object.Then, select a Recipient object, which is the object that receives the recommendation. This would be the Work Order object.And then, choose the object that stores the past interactions between those objects. This would be the Products Consumed object.



Products consumed object

3. Name your recommendation so you can easily identify it later, and give it a brief description.

4. Build the recommendation! This may take up to several hours depending on the amount of data used in your recommendation.

5. When your recommendation is ready, you can deploy it using Einstein Next Best Action.

6. Create a strategy in Strategy Builder, and choose the object you’ll use to display your recommendations.

Then, add your new recommendation to the strategy and run it.

7. Go back to the Work Order record page, and edit the page in App Builder.

8. Drag the Einstein Next Best Action component to the page, add your new Action Strategy, and click Save.

9. And that’s it! We can now see the product parts most likely needed for the work order.

After a user accepts or rejects a recommendation, a flow will execute.

Best practices for admins when building recommendations

Before deploying recommendations, admins can configure the settings of their recommendation to improve the quality and performance. Here are some best practices to keep in mind.

Use segments
Segment your Recipient or Recommended Items objects to focus only on relevant records.

Exclude irrelevant fields
By default, Einstein considers all the fields in the Recipient and Recommended Items objects. You can exclude fields that aren’t relevant to your recommendation. Doing so can improve performance and mitigate some kinds of bias.

Define positive and negative interactions
The way you define positive and negative interactions can affect your recommendation’s performance. You can get better results if you define a positive interaction as the desired outcome. An example of a positive interaction is when a contact purchases a product.

Negative examples aren’t required but can help improve the recommendation. They give useful predictive signals, when available. An example of a negative interaction is when a prospect explicitly rejects a promotion.

Keep in mind, building a good recommendation that meets your business needs is an iterative process. If you’re not satisfied with the quality of your recommendation, you can always refine it.

How do I get started?

Einstein Recommendation Builder is available to existing Lightning Platform Plus and Service Cloud Einstein customers. In April, it will be available to Next Best Action Additional Requests customers.

To dive deeper into Einstein Recommendation Builder, we recommend the Help & Training documentation.

Watch the Trailhead LIVE episode Admin Best Practices: Introducing Einstein Recommendation Builder, and make sure to tune in to the Einstein Recommendation Builder Demo.

Be sure to engage and share feedback or even use cases with us in the Salesforce Einstein Group on the Trailblazer Community. We’d love to hear from you!

AI for Admins: What You Need to Know to Make Einstein Bots a Success

Einstein Bots interact with your customers quickly and accurately with automation and artificial intelligence (AI) capability. In terms of AI initiatives, the biggest challenge is measuring business impact. One reason most AI projects fail is that people are looking at “model performance” instead of business value, such as how much money, in either additional revenue […]

Einstein standing next to text that says, "AI for Admins: Einstein Article Recommendations in Flow."

AI for Admins: How to Deflect Cases with Recommended Articles in Flow

What if you could make that long queue of password reset emails and return policy questions disappear—saving time for your agents to focus on more complex cases? Einstein Article Recommendations supercharge agent productivity by recommending relevant articles for a given case in the service console. Starting in Summer ’22, you can invoke your Einstein Article […]

Mountains and greenery next to text that says "Get Better Natural Language Search Results."

How to Get Better Natural Language Search Results

We’re very excited to announce that we’re enhancing the natural language features as part of Einstein Search. As of Summer 2021, Einstein Search, which includes personalization, actionable results, and natural language capabilities, has been the default search experience for customers within the Lightning Experience. More specifically, Natural Language Search allows users to type queries into […]


Have an Idea for a Story?

We are all about the community and sharing ideas.
Do you have an interesting idea or useful tip that you want to share?