HAS AI ROI GONE AWOL?
I recently had coffee with a former colleague who's leading their company's AI transformation. When I asked how it was going, they paused and said, "Honestly? We're deployed multiple models and built a few inhouse tools; everyone's excited, but I have no idea how to show if it's making a difference… I can't show any ROI."
I've been thinking about that conversation ever since.
Throughout my career, I've seen more than a few technology initiatives follow a familiar pattern: early enthusiasm, scattered adoption, then a frantic effort to justify the investment when the CFO starts asking questions. AI is no different—except the questions are coming quicker, and the stakes feel significantly higher.
The problem isn't that AI doesn't create value. It just needs to be measured differently.
Most companies treat AI ROI as a binary question: "Did it impact my bottom line?" But that approach misses how AI actually creates value in organizations. You can't measure everything the same way, and if you try, you'll either overinvest in the wrong places or miss the opportunity entirely.
After working through several AI initiatives across multiple departments, I've come to see AI value in three distinct categories:
Personal Productivity – Individuals using AI to work faster and smarter
Operational Efficiency – Teams incorporating AI into processes to eliminate bottlenecks
Revenue Growth – Products leveraging AI to drive incremental top-line revenue
Each category requires a completely different approach to investment, measurement, and risk. Each builds on the one before it.
The secret to success: Sequence matters.
Organizations that try to jump straight to revenue-facing AI without building capability and confidence in the earlier phases tend to stumble. Meanwhile, teams that treat personal productivity as "pilots" miss the insights that make operational efficiency possible.
Over the next two posts, I'll break down how value actually builds across these categories and share a practical framework for optimizing value and measuring results.
If you've been wrestling with how to make AI investments defensible and sustainable, this series is for you.
What's your biggest challenge with AI ROI right now? I'd love to hear what's working (or not) in your organization.
[Part 1 of 3]
