Several sellers on a mid-market team stop acting on the AI-generated next-best-action suggestions that appear in their Dynamics 365 Sales workspace. In interviews, sellers say the suggestions feel arbitrary because they cannot tell what prompted a given recommendation. A consultant is asked to improve adoption without removing the feature.
Select an answer to reveal the explanation.
Short Explanation
When people stop trusting a recommendation because they can't see the reasoning behind it, the fix is to show them the reasoning, not to just show them more recommendations or fewer of them. Sellers here aren't complaining about too many or too few suggestions, they're complaining that each one feels like it fell out of a black box. So the move is to expose the actual signals driving a given nudge, things like a gap in recent activity or a cooling engagement trend, so the seller can see the logic and decide for themselves whether to act. Turning up the refresh rate just means more black-box suggestions arriving faster, which makes the trust problem worse, not better. Narrowing the suggestions to bigger deals changes which opportunities get flagged but still hides the why. And swapping the whole feature out for a script written by a manager solves the trust issue only by throwing away the personalized, data-driven part entirely. The real fix is transparency, not volume, scope, or replacement.
Full Explanation
The correct answer is A. The sellers' complaint is specifically about trust and transparency, not volume or scope, so surfacing the rationale behind each recommendation, the actual signals like an activity gap or a slipping engagement trend that triggered it, directly addresses why they are ignoring the feature. Option B is incorrect because refreshing suggestions more often increases how frequently an unexplained recommendation appears without making any single one more understandable, which would likely worsen the frustration. Option C is incorrect because narrowing suggestions to high-value opportunities changes which deals get recommendations but does nothing to explain the reasoning behind them, leaving the core trust problem untouched. Option D is incorrect because replacing the AI feature with a static script abandons personalized, data-driven guidance entirely, which solves the adoption problem only by eliminating the capability the team was asked to improve, not by fixing it. For a consultant, the key insight is that adoption problems rooted in a lack of explainability are solved by adding transparency into the AI's reasoning, not by adjusting the feature's frequency, scope, or replacing it with a non-AI alternative.