Meet Einstein Discovery
Learning Objectives
After completing this unit, you’ll be able to:
- Describe the range of Einstein Discovery features and capabilities.
- Explain key Einstein Discovery concepts including model metrics, insights, predictions, and improvements.
What Is Einstein Discovery?
Salesforce Einstein Discovery augments your business intelligence with statistical modeling and supervised machine learning in a no-code-required, rapid-iteration environment.
Einstein Discovery enables you to:
- Identify, surface, and visualize insights into your business data.
- Predict future outcomes and suggest ways to improve predicted outcomes in your workflows.
Note: Einstein Discovery requires either the CRM Analytics Plus license or Einstein Predictions license, both of which are available for an extra cost.
Target Business Outcomes to Improve
Begin by selecting a business problem you want to solve, typically monitored as a key performance indicator (KPI). Einstein Discovery-powered solutions address these use cases:
- Regression for numeric outcomes represented as quantitative data (measures), such as currency, counts, or any other quantity.
- Binary classification for text outcomes with only two possible results. These are typically yes or no questions that are expressed in business terms, such as churned or not churned, opportunity won or lost, employee retained or not retained, and so on.
- Multiclass classification for text outcomes with 3 to 10 possible results. For example, a manufacturer can predict, based on customer attributes, which of five service contracts a customer is most likely to choose.
Wrangle Data to Analyze
With your business outcome in mind, you prepare the data that Einstein Discovery uses to analyze for insights and train your model. Einstein Discovery can analyze millions of rows and many columns of data! And Einstein Discovery can help you select which columns are the most correlated to the outcome you want to improve. Engage the data integration power of CRM Analytics to load and transform data from one or more data sources into a CRM Analytics dataset. Pull data from Salesforce and external sources. Supplement data cleansing and wrangling tasks with third-party tools and utilities. Effective data prep drives successful Einstein Discovery solutions.
Create Predictive Models Based on Your Data
An Einstein Discovery model is a collection of performance metrics, settings, predictions, and data insights. Einstein Discovery walks you through the steps to create a model based on the outcome you want to improve (your model's goal) and the data you’ve assembled for that purpose (in the CRM Analytics dataset). No coding or machine learning knowledge required!
Use performance metrics to evaluate how well the model predicts future outcomes. The metrics are based on the type of model, such as binary classification, regression, or multiclass classification) for the outcome variable in your model.
Gain Insights into Your Data
Einstein Discovery uses the model settings to analyze your data and produce insights. Data Insights reveal statistically significant findings in your data, and are presented as rich visualizations with natural language explanations that are easy to understand.
Insights provide a starting point for you to investigate trends, patterns, and correlations in your data. Insights tell you what happened (descriptive insights), why it happened (diagnostic insights), and how different variables compare (comparative insights). Then, you can use the model to simulate interactive, what-if scenarios for what could happen (predictions) and ways to improve predicted outcomes (improvements).
Generate Predictions and Improvements
A model is the sophisticated, custom mathematical construct that Einstein Discovery creates based on the settings and the data. Models are based on a comprehensive, statistical understanding of past outcomes that are used to predict future outcomes and suggest improvements. Einstein Discovery supports rapid model development, deployment, and automated maintenance.
You can explore your model's predictions as what-if scenarios: select a variable (or combination of variables) to predict outcomes. Select an actionable variable to see ways to improve predicted outcomes.
A prediction is a derived value, produced by a model, that represents a possible future outcome. Predictors are the variables that contribute to the predicted outcome. Top predictors contribute most significantly.
An improvement is a suggested action that a user can take to improve the predicted outcome. Improvements are associated with actionable variables which users can possibly control, such as the shipping method or a subscriber’s membership level.
Improve Insights and Models with Data Alerts
Salesforce notifies you when it detects issues in your data, such as outliers, missing values, or dominant values.
Bias detection helps you produce ethical and accountable insights and models. Enable bias detection to flag data that could potentially be associated with unfair treatment for sensitive variables, such as race, gender, religion, national origin, sexual orientation, disability, or age. Einstein Discovery detects disparate impact and proxy variables that are highly correlated with sensitive variables.
Operationalize Models into Salesforce
Einstein Discovery uses models to predict outcomes and suggest improvements. You can have Einstein Discovery build your model, or you can upload externally built models. To ensure the best quality model, check its quality metrics and alerts and tune it accordingly. When it’s ready to go, Einstein Discovery walks you through the steps to deploy your model into Salesforce.
Get Predictions and Improvements Everywhere
Once deployed, use Einstein Discovery models to get predicted outcomes and suggested improvements with clicks-not-code in:
- Lightning record pages
- Experience Cloud sites pages
- CRM Analytics Data Prep recipes and dataflows
- PREDICT function used in process automation formulas
- Salesforce flows (with Flow Builder)
- Tableau flows, dashboards, and calculated fields
With Einstein Prediction Services and the Insights API, you can programmatically interact with Einstein Discovery-powered insights, models, and predictions via REST or Apex API calls.
Augment Tableau with Einstein Discovery Predictions and Improvements
Einstein Discovery in Tableau brings trusted, real-time predictions and improvements to your Tableau data in a no-code-required, rapid-iteration environment.
Inject predictions and improvements directly in Tableau dashboards, calculated fields, and Tableau flows. To learn more, see Get Predictions in Tableau.
Jumpstart Solutions with Model Templates
Einstein Discovery model templates provide out-of-the box, fast-start solution kits that you can customize to your solution requirements. Model templates provide end-to-end implementations of common business use cases.
Einstein Discovery handles the heavy lifting of building the initial training dataset, generating the Einstein Discovery model, and deploying the model into Salesforce to get predictions on your data. Designed to integrate seamlessly with your Salesforce data, model templates are built with industry best practices in mind.
Discover Report Insights
Einstein Discovery for Reports scans your report data—quickly and thoroughly—using artificial intelligence and comprehensive statistical analysis.
Einstein Discovery for Reports goes deep into the report data, explores underlying patterns, identifies insights, and surfaces those insights with correlations, charts, and explanations that are easy to understand. To learn more, check out Einstein Discovery for Reports: Quick Look.
Where to Go From Here
Now that you’ve learned about all the functionality that Einstein Discovery has to offer, consider the following resources to continue your learning journey into Einstein Discovery-powered insights, predictions, and improvements.
Resources
- Salesforce Help: Explain, Predict, and Take Action with Einstein Discovery
- Trailhead: Gain Insight and Improve Outcomes with Einstein Discovery
- Trailhead: Ethical Model Development with Einstein Discovery: Quick Look
- Trailhead: Responsible Creation of Artificial Intelligence
- Quip: Popular Learning Resources
- Technical Paper: Understanding the Differentiating Capabilities and Unique Features of Salesforce Einstein Discovery in the Machine Learning Space