A sales manager reviews a recorded call where the customer expressed frustration about a delayed shipment and hinted at evaluating a competitor. Conversation intelligence transcribed the call accurately, but the opportunity was not flagged as at-risk and no alert appeared in the manager's dashboard. The manager asks a consultant why. What is the most likely cause and correct fix?
Select an answer to reveal the explanation.
Short Explanation
The system heard the frustration and heard the competitor mention just fine — the transcript proves that. What it didn't do is connect what it heard to a rule that says flag this. Sentiment and risk detection only fires when someone has told it what to listen for: which keywords matter, which topics count as risk, how negative is negative enough. If nobody configured a trigger that matches a customer being unhappy about a delay and name-dropping a competitor, the call sails through analyzed but silent. That's a configuration gap, not a capability gap — the fix is teaching the system what a red flag sounds like for this business, not waiting for a longer recording, not making a manager manually flag things by hand first, and not chasing a consent technicality that has nothing to do with whether an already-processed call can raise an alert.
Full Explanation
The correct answer is A. Conversation intelligence relies on configured sentiment thresholds, keyword tracking, and topic detection to decide when a call's content should generate a risk flag; if the frustration and competitor mention did not trip any configured trigger, the detection rules themselves are the gap, and adjusting thresholds or adding relevant keywords and topics is the direct fix. Option B is incorrect because a minimum call-duration setting governs whether a call is analyzed at all, and the scenario states the call was transcribed and analyzed accurately, so duration was not the blocking factor. Option C is incorrect because flags are meant to be generated automatically from detected signals; requiring a manager to manually mark an opportunity at-risk before any alert can appear defeats the purpose of automated conversation intelligence. Option D is incorrect because there is no separate consent field gating whether a detected signal produces a flag; consent and recording policies govern whether a call can be captured and analyzed at all, not whether an already-analyzed transcript can trigger alerts.