The data for the Einstein Lead Scoring engine is the historical leads. Einstein looks into each of the fields and attributes of the historical lead and then finds what led to the lead conversion. The correlation between the new lead attributes and the historically converted leads defines the lead score.

Einstein Lead Scoring uses the existing fields on the Lead record to find data points that have been responsible for the lead conversion. The Einstein Lead Scoring engine automatically adopts the best model to score leads.

To measure the accuracy, Salesforce provides a few wave dashboards and reports out of the box.

Since we do not have access to the free developer organization for the Einstein Lead Scoring feature, the information presented is drawn from the Salesforce Help and Training document. We will use some screenshots from the Help and Training document (https://help.salesforce.com/articleView?id=einstein_sales_lead_insights_parent.htm&type=0&language=en_US&release=208.8) to understand how these native Wave Analytics reports can help measure accuracy.