Saurabh's Scoop │ July 22, 2025

Welcome to Saurabh’s Scoop, where I bring you weekly insights from HFS Research and my personal reflections on the top stories shaping the intersection of technology and business.

This week, let’s talk about the AI hype machine—and why so little of it delivers.

GenAI and Agentic AI are stuck in “death by a thousand POCs.” Despite billions poured into pilots, prototypes, and vendor demos, only 10–15% of AI initiatives ever make it to production.

 At HFS, we’ve consistently spotlighted the foundational issues holding enterprises back—what we call the four big organizational debts:

• Process debt: Broken or inconsistent workflows

• Data debt: Poor quality, fragmented, or inaccessible data

• Skills debt: Lack of AI fluency and change readiness

• Tech debt: Legacy systems sprawl that can’t scale AI

But there’s another critical issue enterprises keep missing: We’re writing the wrong business cases!

Too many AI investments are justified by a narrow obsession with productivity and cost take-out—while ignoring real strategic levers that drive value at scale: Performance, Personalization, and Prediction.

Chart of the Week: The 4Ps of AI Value Creation – Performance, Personalization, Prediction, and Productivity   

Over 80% of AI in production today drives value through Performance, Personalization, and Prediction. In contrast, most POCs and pilots remain stuck chasing Productivity.

Yes, productivity is part of the AI value equation—but over-indexing on it is a fool’s errand.

Note: This analysis is based on our upcoming “HFS AI Ideas Book” (available soon on hfsresearch.com) that compiles 1000s of GenAI and Agentic AI use cases by industry, function, stage of deployment, and outcomes. Kudos to my colleagues Dana Daher, Suhas AR, and Niti Juhnjuhnwala for leading this initiative.

Productivity Obsession = Value Suppression

We’ve made productivity the hero of most AI pilots—obsessed with FTE reductions, automation metrics, and operational cost take-out. But when you build your AI business case solely on cost take-out and task automation, you not only limit impact, you trigger resistance. Productivity is part of the story—but not the story.

• AI eats tasks, not jobs. AI mostly automates fragments of roles—not entire positions. These “fractional FTE” gains are hard to realize unless you have a plan to reintegrate or upscale that freed-up capacity.

• Fear erodes adoption. People aren’t excited about AI when it’s sold as a job killer. They resist change. They delay adoption. AI becomes something done to them—not with them. And that’s where value dies.

• Productivity without purpose is a dead end. Time savings don’t equal business value unless you redeploy talent toward innovation, experience, or growth. Without that reinvestment, AI becomes just another sunk cost.

The Big Picture: Start Using a Balanced Scorecard Mapped to the 4Ps of AI Value Creation

To turn AI ambitions into reality, we need to reframe value creation using the 4Ps. Use a balanced scorecard for your AI business case:

If you're stuck trying to justify AI with only productivity metrics, you’re building a business case for a tool—not a transformation.

Shift the narrative. Use the 4 Ps. Build smarter.

This newsletter is inspired by the bold HFS 2030 Vision for Services-as-Software—a transformative outlook on how technology will redefine the way enterprises operate. In each edition of Saurabh’s Scoop, I’ll connect the stories, trends, and innovations of today to this exciting vision of tomorrow.

Contact me with feedback and suggestions:

Saurabh Gupta, President, HFS Research

• Email: Saurabh.Gupta@HFSResearch.com

• LinkedIn: linkedin.com/in/hfssaurabh