Back to the Blueprint: Why the Original AI Scorecard Still Matters Today

We’re surrounded by noise. AI is everywhere. Layoffs are up, funding is weird, pilots are collapsing, and people are still pretending that duct-taping APIs together is a strategy.

But here’s the thing—we’ve seen this before. And we had a framework for it.

Before the hype cycles and AI demo clips, there was the AI First Scorecard—a blueprint to assess whether an organization was truly built for AI or just playing with it.

Why This Still Hits Home

Every company wants to say they’re “doing AI.” But most can’t move past pilot. Why?

Because adoption doesn’t equal architecture. Strategy doesn’t mean execution. And speed without structure is just chaos.

We keep hearing “if you’re not using AI, you’re a zombie company.” That kind of noise pressures teams to chase tools instead of solving problems. Meanwhile, Big Tech is sending conflicting signals, and frameworks like NIST (which I talked about last week) don’t offer much for those actually building.

Then there's Publicis Sapient’s take. Their AI Scorecard breaks it down into something practical: Foundational → Emerging → Developing → Optimized It’s not flashy. It’s useful.

They focus on two things:

• Readiness – do you even have your house in order?
• Maturity – are you doing anything meaningful with it?

Simple, but effective. And what I respect most—they call out that tech alone isn’t the point. Purpose and culture matter.

It’s Not New—It’s Transferable

I also came across MDPI’s revisit of the original Balanced Scorecard (Kaplan & Norton, 1992). It measured success across financials, customers, internal ops, and learning. That structure changed how companies thought about execution.

We’re in the same place now with AI. The problem space shifted, but the need for multidimensional alignment didn’t.

These ideas aren’t new. But the discipline they require? That’s still rare.

What the Original Scorecard Actually Measured

This wasn’t a checklist. It was a stress test.

• Tech Stack – are your systems modular or a spaghetti mess?
• Architecture – can teams build on top of each other without breaking everything?
• Teams – do the right people actually own the work?
• Integration – does AI touch real workflows or just sit in a slide deck?
• Governance – is trust built in, or patched on later?
• Feedback – do you learn in real time?

It didn’t just measure outputs. It connected tech to outcomes:

• Business value
• Efficiency
• Competitiveness
• Societal impact (yes, that too)

4 Levers That Still Matter

1. Reference architecture – start from something solid
2. Thin crossing points – don’t make systems dependent on each other
3. Empowered teams – let the people closest to the work solve the work
4. Value loops – close the loop between users, data, and decisions

Publicis Sapient basically said the same thing: align purpose, insights, governance, and learning. No shortcuts.

So Where Are We, Really?

Everyone keeps asking: “Is this the singularity?”

No. Most AI projects today are… empty. Startups raise on hype. Enterprises race to deploy without infrastructure. And governance? Usually an afterthought.

We’re seeing:

• A brutal job market
• Failed implementations
• A lot of performative “AI” with no staying power

What to Actually Do

• Reassess what you’re building. Is it architectural—or duct tape?
• Score your AI maturity. Not with a vendor demo—use real benchmarks.
• Evaluate your models, compute needs, and data strategy
• Implement sovereignty and governance now—not after your launch
• Recommit to alignment. It beats speed every time.

If you’re building, investing, or regulating—step back. Look at the architecture.

Go back to the blueprint.

Let’s stop pretending this is new. Let’s build AI systems that work.

References:

https://www.publicissapient.com/insights/introducing-the-ai-scorecard#:~:text=Introducing%20the%20AI%20Scorecard:%20The,of%20AI%20readiness%20and%20maturity.

https://www.mdpi.com/2673-8392/5/1/39