ROI Analysis is More Important Than Ever
ROI Analysis is More Important Than Ever
Three background facts:
- Lots of companies are worried about the cost of AI tokens and it is causing some to scale back to save costs. They seem to do this by rationing their departments to use less.
- A Feb 2026 study by McKinsey says that approximately 70% of companies implementing AI get no increase in EBIT. 30% do but they did not report the typical magnitude of the EBIT across that percentage. There are some respectable ones out there.
- McKinsey also concluded that for every dollar spent on an AI project at least $3 needs to be spent on change management to get to a successful conclusion.
This is an improvement on 2025 AI projects where 90% were reported as failures.
With 67% of supply chain digital investments now allocated to artificial intelligence, more than half of chief supply chain officers (CSCOs) are uncertain about the ROI, according to a recent Gartner survey.
Gartner surveyed 394 supply chain professionals from organizations with annual revenue of at least $250 million, conducted between November 2025 and February 2026, to understand how their organizations allocate digital investments. Gartner also surveyed 125 senior supply chain leaders from January through April 2026 to gather AI use cases and analyze ROI. These findings show that even if organizations are investing heavily in AI, measuring the business impact may still be challenging.
Too many AI projects fail because they are managed by senior people who no longer know how the detail of the existing process works on the ground. AI solutions get developed to improve the process documented in the company process manuals. But the people in the field found all kinds of work arounds to make the documented processes work. They never got written down.
Now you have the background learn more about the structure of the ROI model and how it is used.