How do I set up attributes and filters for win probability and lead scoring prediction?
You can configure which fields and records that you want AI models to consider, to predict opportunity win probability or lead score.
You can do one of these actions on the Opportunity Win Probability and Lead Scoring pages:
- Use filters to exclude records for parent objects such as Lead, Opportunity, or Account. For example, for the Opportunity object, you can set a filter as Sales Channel Equals Indirect. This means that if an opportunity has the value for Sales Channel as Indirect, it won't be considered for prediction.
- Replace ready-to-use fields (also called attributes) for any object used by AI
models with custom fields that have matching data types.
For example, for an opportunity win probability prediction in the Activity object, you might not want the models to consider the
ActivityTypeCode
field. You might instead want the model to consider data from theActivityCategory_c
custom field. When you select this replacement, the models will consider activity category across opportunities that were won as one of the factors to predict the win probability of new opportunities.
- In Setup and Maintenance, go to:
- Offering: Sales
- Functional Area: Intelligence in Sales
- Task: Configure AI Apps in Sales
- On the Intelligence in Sales page, click Configure AI Apps in Sales.
- On the Configure AI in Sales page, select one of these options based on what you
want to configure:
- Opportunity Win Probability: Select this to configure attributes for opportunity win probability prediction.
- Lead Scoring: Select this to configure attributes for lead score prediction.
- On the Opportunity Win Probability or Lead Scoring page, from the Object list,
select the object that you want to remove or replace attributes for.Note: The Object list consists of parent and child objects. For parent objects, you can also apply a filter to select specific records in objects that you want to exclude in the model.After you select an object, attributes related to that object are listed in the Model Feature Configuration table. In the Model Feature Configuration table, you can do one of these:
- Click the Edit icon to replace a ready-to-use
attribute with a custom attribute. Select a custom attribute from
the list in the Replace With field. You can only replace an
attribute with another if they share the same data type. For
example, if you’re replacing an attribute whose data type is string,
the attribute you’re replacing it with must also have string as the
data type.Note: You can only replace some ready-to-use attributes.
- Click the Delete icon to delete a custom attribute mapping for a ready-to-use attribute. Deleting this mapping clears any replacement selections that you made for the ready-to-use attribute in the Replace With field.
- Click the Edit icon to replace a ready-to-use
attribute with a custom attribute. Select a custom attribute from
the list in the Replace With field. You can only replace an
attribute with another if they share the same data type. For
example, if you’re replacing an attribute whose data type is string,
the attribute you’re replacing it with must also have string as the
data type.
- Click Save.
The models will now use the dataset from the attributes based on how you've set them up here.