A consultant is preparing to enable AI-driven opportunity scoring for a sales organization that has just migrated to Dynamics 365 Sales from a legacy CRM. Before turning the feature on, the consultant checks the environment against Microsoft's stated prerequisites. (Select TWO.) Which two conditions must be met for the scoring model to function reliably?
Select all correct answers, then click Submit.
Short Explanation
Two different things have to be true before scoring will actually work for people. First, the model needs enough history to learn from — and it needs the full picture, wins and losses both, not just the deals that closed successfully, because migrating from an old CRM is exactly the kind of moment where that history can go missing or get left behind. Second, having the model running in the background doesn't automatically mean anyone can see its output — access to predictive scoring is gated by role and license, so a seller without the right assignment sees nothing no matter how good the model is. What doesn't belong on this list: making everyone hand-fill a probability field before scoring can even start, which is backwards, since generating that kind of estimate automatically is the entire point. And there's no real tension between running sequences and running scoring — nothing about one requires switching the other off.
Full Explanation
The correct answers are B and C. AI-driven opportunity scoring learns patterns from historical outcomes, so it needs a sufficient volume of closed opportunities, spanning both wins and losses, to identify what separates a deal likely to close from one likely to stall; without that history the model has nothing reliable to learn from, which is a real risk right after a migration from a legacy CRM where that history may not have carried over cleanly. Separately, predictive scoring is a feature gated by security role and license assignment, so sellers who lack the appropriate role or license will not see scoring even once the model itself is functioning. Option D is incorrect because opportunity scoring is generated by the model itself and does not require every open opportunity to already carry a manually assigned probability-to-close percentage as a precondition; requiring that would defeat the purpose of automated scoring. Option A is incorrect because Sales Accelerator sequences and opportunity scoring operate independently and are not in conflict, so there is no reason to disable one to enable the other, and no such prerequisite exists in the product.