Everyday AI Podcast – An AI and ChatGPT Podcast Ep 725: Measuring AI ROI: Why you’re doing it wrong and the 7 Steps to fix it (Start Here Series Vol 11)
98 snips
Mar 3, 2026 They tackle why most companies fail to show GenAI ROI and how using old transformation playbooks breaks implementations. The conversation highlights benchmarks where models match experts, critiques flawed negative studies, and points to quantitative research showing positive returns. They outline hidden productivity losses, a simple ROI formula, and a seven-step blueprint with testing, baseline assessment, and ongoing retesting.
AI Snips
Chapters
Transcript
Episode notes
ROI Debate Is Missing The Point
- The ROI debate about AI is largely pointless because models already match or beat experts and run vastly faster.
- Jordan Wilson cites OpenAI's GDPVal showing top models tie/win 70% of tasks and often complete them ~100x faster.
Bad Studies Can Skew Market Perception
- Viral negative headlines like MIT's '95% zero ROI' came from 52 qualitative interviews and lacked quantitative backing.
- Jordan says that study functioned more like marketing and misled markets despite broad media pickup.
Large Studies Show Positive AI Returns
- Multiple large quantitative surveys show positive AI ROI, e.g., IDC $3.70 per $1, Wharton and Google Cloud ~74%, and Deloitte 84%.
- Jordan uses these studies to counter the narrative that AI pilots usually produce no ROI.