Fabrikam’s CFO argues that the only ROI metric that really matters for generative AI is “cost per 1,000 tokens,” because that is how Azure OpenAI is…

The statement is false. While "cost per 1,000 tokens" is a critical operational metric for managing Azure OpenAI expenses, it is not a measure of Return on Investment (ROI). True ROI evaluation must connect AI usage to tangible business outcomes. This includes metrics like hours saved through automation, increased revenue from AI-driven insights, improved customer satisfaction scores, or reduced operational costs in other areas. Focusing solely on token cost ignores the "return" component and leads to poor strategic decision-making about AI investments.

This is a True/False question, so there are no other options to evaluate. The provided answer 'B' corresponds to 'False'.

1. Microsoft AI Business School, "Define a business strategy for AI". This courseware emphasizes the importance of identifying clear business problems and measuring success with key performance indicators (KPIs) tied to business outcomes, not just technical costs. (See Module: "Create a holistic AI strategy").

2. Microsoft Learn, "Plan and manage costs for Azure OpenAI Service". While this document explains the token-based pricing model, it also advises on monitoring usage in the context of the application's purpose, implicitly connecting cost to the value being delivered. The goal is cost optimization for a valuable service, not just cost reduction in isolation.