Create scoring models
With Customer Insights - Journeys, you can use a powerful lead-scoring builder to define your scoring criteria and model more efficiently.
Dynamics 365 Customer Insights - Journeys scores leads by using demographic attributes and behavioral interactions. An entity who meets your target demographic profile and interacts with your marketing materials receives a high score. Interacting with marketing materials includes activities such as opening emails, downloading content, browsing your website, or attending an event.
Scores help identify the best customers, but not everyone in your organization might easily differentiate between a good or poor score. You can use grades as a way to offer an interpretation of the score (such as cold, warm, or hot).
Turn on the feature
To use scoring models, an administrator needs to turn on the feature in the settings by following these steps:
Go to Settings > Overview > Feature switches.
Turn on the Lead Management feature and then select Save in the upper-right corner.
After you turn on the feature, a new feature appears on your Customer Insights - Journeys navigation bar at Lead management > Scoring models.
Create a new scoring model
To create a new scoring model, follow these steps:
Go to the Real-time journeys work area within the Marketing app.
Go to Lead Management > Scoring Models.
To start creating a new scoring model, select the New button.
Provide a name for the model.
After naming the model, you can specify conditions to target a specific audience. This specification helps you avoid generating extra scores that you don't need. (You can add or adjust these conditions after you're in the scoring model builder.)
When you're ready to start building your model, select the Create button.
In the model builder experience, on the right side of the interface, consider the following data to start building your conditions:
Attributes - Refer to demographic information that comes from the entity that's being scored or related tables to the entity, such as contacts or accounts.
Behavioral - Interactions represent engagements or activities.
To create a new condition, start by selecting an attribute or interaction.
For attribute conditions, you add the attribute, choose the operator, and define the value(s) for the condition.
You can also build more complex queries that reference other tables (such as Event Registration or Account) to improve your scoring model. When adding an attribute that relates to a different table, you can search for the attribute, define how the two tables are related, and then use the related table in your scoring model.
For behavioral conditions, after you add the behavior, you can select Edit to further define.
Choose whether the lead has triggered or has not triggered the interaction, the number of occurrences, and the time window in which the interaction occurred (for example, the customer clicked on an email once in the last 30 days). When you finish identifying the interaction details, select Done.
Further refine the condition by adding interaction metadata to the condition. For example, if you choose the marketing form-submitted interaction, you could identify a specific form rather than scoring on all form submissions.
After you add the condition, you can define how many points a customer gets by fulfilling this condition. Points can be negative or positive.
Use the dropdown menu above the condition to change between Add and Subtract.
Manually enter the point value or use the up or down arrows to adjust.
Continue adding new groups by selecting the Create new group option.
Add grades to your scoring model (optional).
If you want to help communicate the differentiation between a good or poor score, you can use grades in your model.
Go to the second option on the vertical bar to the right of the window and then select Add grade.
Enter a grade title (such as Hot or Cold).
Select the operator (such as less than, more than, or between).
Define the score that applies to that grade. If you select is between as the operator, you need to enter a range, such as 0 to 12 points.
Continue adding grades as desired.
Review the settings of the scoring model.
The third option in the vertical bar to the right of the window allows you to access the model settings.
On this page, you can visualize the target audience that's getting scored, modify the conditions that you initially defined for the model, and then view the Business Unit that the model is scoring (in case it applies). If you're working in a business unit, then the model scores only the leads from that business unit. This option doesn't appear unless you have business unit scoping turned on.
Additionally, you can decide if you want to apply parent contact interactions or lead interactions.
You can store interactions at the lead level if you market to leads or at the contact level if you market to contacts. Customer Insights - Journeys scoring models empower you to decide whether the scoring model should account for parent contact interactions.
If you want to include parent interactions, select the Count parent contact interactions checkbox.
Publish the scoring model.
When your model is ready, select Publish to start calculating scores.
You can still view the conditions, but you can't modify any of them. Depending on the volume of leads and the complexity of the model, the amount of time to calculate scores varies.
Note
The scoring model initially targets only lead entities. In the future, the feature may support other entities. Watch for future updates and enhancements to the scoring model capabilities.
Note
You can only use Microsoft Dynamics 365 standard interactions to create conditions.
For more information, see Create scoring models.
Review key insights and scores for each lead
After you publish your lead-scoring model and calculate scores, you can select the Insights tab in the scoring model to review the following components:
Total number of leads scored
Score for top 10% of leads
Score distribution among all leads
Volume of leads for each grade
This data helps you identify high-potential customers based on scores. You can also use data as input for other orchestration features. For example, you might want to create a personalized experience for warm leads to turn them into hot leads. Start by creating a segment for leads with a score that's equal to the warm range that's identified in the scoring model. Then, you can build a journey to send curated content to those leads.
If you want to review more granular results, you can select the Scores tab in the scoring model. The Scores tab includes all leads that the model scores with their respective scores and grades. If you want to validate your scoring models' results, you can access each lead record and review the calculated scores.
For more information, see Visualize results from your scoring model.