Troubleshooting Einstein Opportunity Scoring

Simple Demo Orgs (SDOs) do not support any type of AI feature including Opportunity Scoring. You can confirm if an org is a SDO by navigating to Custom Metadata Types from Setup, if you see one with the name "QLabs", it's a SDO.

Initial Troubleshooting Steps:

• Run the Einstein Readiness Assessor and ensure there is a green check mark next to Opportunity Scoring. If there is not, scroll to the Opportunity Scoring section of the report and check what is blocking the feature from being enabled.
• When enabling the feature, verify that the affected user has both the "Sales Cloud Einstein" permission set license, and the "Sales Cloud Einstein" permission set. If they do not, make sure to assign those permissions to the user and try again. If the org does not have those licenses available, reach out to your Account Executive.
• Determine whether or not you have "Opportunity Scoring for Everyone.“ Please see the frequently asked questions section additional details on how to determine if you have access.
• Ensure the org meets the data requirements as outlined below.

If the assessor is being run in a Sandbox, the following requirements must be met:

• Opportunity history shows one update to each closed opportunity on average. If this is not the case, enable Opportunity History to ensure opportunity changes are being captured in the sandbox.
• Using Opportunity Scoring in sandboxes is only supported for a full copy sandbox with all of your production data (at least 180 days). Note: Manipulating data in a sandbox prior to running the Readiness Assessor could cause unexpected results with the Assessor. 

Data Requirements: 

Certain data requirements will need to be met for the feature to function properly. Each of these is checked by the Einstein Readiness Assessor, but there may be situations where you would want to check on these manually (e.g. the Readiness Assessor says you need at least 200 opportunities closed-won in the last 2 years).

• There must be at least 200 closed won opportunities in last 24 months, each with a lifespan of at least 2 days.
• There must be at least 200 closed lost opportunities in last 24 months, each with a lifespan of at least 2 days.

The following SOQL queries will gather an exact number of opportunities closed-won and closed-lost in the last two years. This number must be greater than or equal to 200: 

• SELECT count() FROM Opportunity WHERE IsClosed=true AND IsWon=true AND CreatedDate = LAST_N_DAYS:730
• SELECT count() FROM Opportunity WHERE IsClosed=true AND IsWon=false AND CreatedDate = LAST_N_DAYS:730

Frequently asked questions about Opportunity Scoring:

Q: How do I tell if I have Opportunity Scoring For Everyone available for my org? 

A: In Lightning, navigate to Setup > Einstein > Einstein Sales > Assisted Setup. If there is an option for "Sales Cloud Einstein for Everyone" then you have Opportunity Scoring for Everyone. The “Sales Cloud Einstein for Everyone” permission set is required to verify.

Q: Who is eligible to get Opportunity Scoring for Everyone?

A: In the Spring ’20 release, Salesforce announced that it will make Einstein Opportunity Scoring available to all Sales Cloud customers.

Q: One of the Top Negative Influences for a score says "No Recent Updates to Opportunity." What does that mean, and how is it calculated?

A: Once a model is created, it establishes a threshold number of days where once an opportunity exceeds the threshold without any relevant updates, the "No Recent Updates" top negative will contribute to the score. Fields that count toward an update are Created Date, Close Date, Stage Name, Forecast Category, and Amount. If the opportunity is modified again or edited without changing anything after a while, the modification time filed will be updated and its value will be presented in the key factor. 

Q: My Opportunity scores don't seem to be correct, why?

A: Verify the following:

• Opportunity Scoring is enabled.
• The permission set is assigned.
• 36 hours has passed after the enablement in setup.
• Score field is added to List-view/page-layout.
• Your org has passed the Readiness Assessor.

Q: What fields were used to build my model, and what is the importance of each field?

A: Build a model factors report and review the following:

• Do you use the correct field in the model? Have you tried using a different field in the model?
• Things to keep in mind:
  1. Not all of the fields are statistically significant and should be included in the model.
  2. When hovering on a score, only the top 4 predictive factors are shown. What is shown depends on the actual score. If the score is very high, it'll show the top 4 positive. If the score is very low, it'll show the top 4 negatives. Otherwise, there will be a mix of positive and negative. This is why a field might have an influence on the opportunity, but it’s not significant enough in relation to the other fields.
  3. Only the following custom field types are supported:
     • picklist
     • string
     • reference
     • multipicklist
     • boolean