In an AI project, how should business goals relate to model metrics?

• A. They should be independent.
• B. They should be considered late in development.
• C. They should be aligned—model metrics should reflect business goals.
• D. They should be determined by marketing needs.

Answer: (C)

Connecting how a model performs to the business outcomes the organization cares about is essential. When model metrics reflect business goals, every improvement translates into real value, not just better statistics. For instance, if the goal is to boost revenue, the relevant measures might be conversion rate, average order value, or profit impact, so you evaluate changes by those outcomes rather than a generic accuracy score. If the aim is a better user experience or higher retention, metrics such as engagement, churn reduction, or time-to-value become the true indicators of success. If cost efficiency is the priority, you’d focus on latency, throughput, or cost per prediction. In short, align the metrics with the business definition of success, and let that guide decisions about deployment, thresholds, and retraining.

Choosing metrics independent of business goals can lead to optimizing something that doesn’t matter for value. Delaying alignment until late in development robs you of chances to steer design toward outcomes that matter. Relying solely on marketing needs might tilt toward flashy numbers without considering feasibility, risk, or long-term impact.