A Dynamics 365 Sales administrator has already licensed Dynamics 365 Sales Premium and turned on predictive lead scoring in Sales Insights settings. Three weeks after go-live, almost every lead still shows a blank predictive score. Most leads in the system are still open, with only a handful marked qualified or disqualified so far. What is the most likely reason the scores are blank?

The correct answer is the explanation that the predictive model needs a larger volume of leads with a mix of qualified and disqualified outcomes before it can learn patterns and generate reliable scores. Predictive lead scoring is a machine-learning model, and models need a training set of outcomes to learn from. With only three weeks of history and just a handful of leads marked qualified or disqualified, there is not enough labeled data for the model to distinguish patterns that predict success from patterns that predict failure, so it withholds scores rather than guess. As more leads move to a final outcome over time, the model gains the examples it needs and scores begin appearing. The licensing-tier explanation is incorrect because the scenario states Sales Premium is already licensed, which is the tier that includes lead scoring; no further upgrade is required. The security-role explanation is incorrect because a security role gap, such as missing write access to the Leads entity, would typically block the administrator from configuring or viewing the feature entirely, not cause scores to be blank specifically for records lacking outcome history, and the scenario gives no indication of a permissions error. The support-ticket explanation is incorrect because predictive scoring is a configurable feature the administrator turns on directly in Sales Insights settings; it does not require a Microsoft support ticket to activate per environment. The real constraint here is data maturity, not licensing, permissions, or activation mechanics.