Ep. 47 — The AI Value Chain
Workhelix Podcast · 52:00
Episode Highlights
Why the “70% of AI projects fail” statistic is both true and misleading — and what the real failure mode looks like
The measurement maturity curve: four stages from anecdote to real-time ROI attribution
How Meridian Financial built a measurement framework that paid for itself in 60 days
The organizational change that matters more than any AI tool: making measurement someone's job
What Claire would tell her past self about building AI business cases that survive CFO scrutiny
Transcript Excerpt
HOST
“You've said the measurement problem is structural, not technical. What do you mean by that?”
CLAIRE HOFFMAN
“The tools to measure AI impact exist. What's missing is someone owning the question. When I joined Meridian, our AI initiative had a sponsor, a budget, and a roadmap — but nobody's performance was tied to demonstrating that the investment was returning value. The moment we changed that, measurement happened automatically. People find ways to measure what they're accountable for.”
HOST
“So the measurement gap is really an ownership gap.”
CLAIRE HOFFMAN
“Exactly. And the organizations I see struggling the most are the ones treating AI measurement as a technical problem to be solved by the data team, when it's actually an operating model problem to be owned by the business.”
.png){kind=icon}
