Question: In a few sentences, explain how a team’s sense of what users need for a great product experience can help set a desired error rate for a learning system.

Sample answer: When a team knows what level of mistakes users will tolerate, that experience-based judgment can be translated into a desired error rate. For instance, if a photo-search feature can still feel smooth even when it occasionally returns the wrong result, the acceptable error rate may be higher than for a safety-critical system. If even a small number of errors would frustrate users, the target must be much stricter. In this way, the error rate is tied to the product experience the team wants to deliver.

Key points:

• A product serves users, so user experience matters.
• Intuition about acceptable performance gives a practical benchmark.
• That benchmark can be turned into a concrete desired error rate.
• The target should reflect what users need, not an arbitrary number.

Rubric: Full credit for explaining that intuition about the user experience helps translate product needs into a concrete desired error rate.