Enterprise hits and misses - AI, analytics are metrics - what are we missing? Enterprises confront AI Doomsday narratives, and event season rolls on

The article argues that CIOs should stop treating AI ROI as a simple productivity or model-cost exercise and instead measure the full cost stack: inference, verification, and the business context required for reliable outputs. Strategically, it says enterprises do not need perfect data foundations before starting; IT should prioritize measurable use cases, add a reusable context layer, and use deterministic processes where they are cheaper and more dependable than LLMs.