A consultant preparing to turn on predictive lead scoring for a client for the first time confirms the client already has the required premium licensing in place. (Select TWO.) Which two additional conditions must also be true before the feature can be enabled and produce useful scores?
Select all correct answers, then click Submit.
Short Explanation
Getting this feature running for the first time comes down to two real gates, and it helps to keep them straight from the things that sound plausible but are not actually requirements. The first gate is about who can even reach the configuration screen: turning the feature on is an admin-level task, so someone with the right level of access has to be the one flipping that switch. The second gate is about the raw material the model needs to learn from: it has to look at a meaningful stretch of past leads that already closed, some as wins and some as losses, so it can figure out which patterns actually predicted success. Skip either of those and the rollout stalls or the resulting scores will not be trustworthy. What does not belong on this list is any kind of mandatory training course for sellers, since that has nothing to do with whether scores appear, and there is also no need to shuffle records into some separate AI-only workspace before flipping the switch.
Full Explanation
The correct answers are A and B. Enabling predictive scoring is an administrative configuration task, so it requires someone with the appropriate privilege to open the Sales Insights settings and turn the model on for the relevant table; without that access, the feature simply cannot be activated. It also requires a sufficient body of historical closed leads, including both won and lost outcomes, because the model learns which signal patterns correlate with conversion by studying past results, and a training set skewed toward only wins or too small overall will not produce a reliable model. Option C is incorrect because there is no Viva Learning training prerequisite tied to whether scores display; sellers see scores based on configuration and licensing, not completion of external training modules. Option D is incorrect because predictive scoring runs against the existing Dataverse environment where the Leads table already lives; there is no requirement to migrate records into a separate dedicated environment, and doing so would be unnecessary and disruptive. The two genuine prerequisites are administrative access to configure the feature and adequate historical outcome data for the model to train on.