'Do More With Less' Wins in Enterprise AI

I recently watched Daniela Amodei, President of Anthropic, discuss their approach to AI development in a CNBC interview. One phrase stopped me cold: "maximum capability per dollar of compute." Not because it was novel—but because it crystallized a principle I've practiced for 38 years.

In enterprise technology, we're conditioned to believe scale is destiny. Bigger models. More parameters. Larger training runs. The assumption is that whoever spends the most wins. But Amodei articulated something different: the future belongs to those who deliver the most capability per unit of investment.

She's right. And the enterprises quietly winning in AI adoption understand why.

The Efficiency Equation

In my book, "Do The Right Thing in Business Improvement," I formalized what I call the Core Business

Model principle:

Productive Performance = (Functional Capability × Efficiency) / Investment

This isn't abstract theory. It's a framework I've used across 75+ enterprise transformations to help clients understand where to focus for maximum impact. The equation reveals something critical: you can increase productive performance in three ways—boost functional capability, improve efficiency, or reduce investment. Most companies obsess over the first while ignoring the other two.

The "bigger is better" approach focuses solely on functional capability: more compute, larger models, additional parameters. But this strategy has diminishing returns and escalating costs. Meanwhile, efficiency improvements and smarter investment allocation often deliver superior results at fraction of the cost.

Enterprise buyers increasingly understand this math. They're not asking "which model has the most parameters?" They're asking "which approach delivers reliable results we can afford to scale across our operations?"

Why Efficiency Wins in Enterprise Reality

I've spent decades translating complex technical innovation into business adoption. The pattern is consistent: technical superiority doesn't automatically translate to market success. Three factors determine enterprise AI adoption, and all favor efficiency-driven approaches.

1. Cost Predictability Matters More Than Raw Capability

At KeyBank, I led development of an AI-like automated loan decisioning system before we called it AI. We had options: build the most sophisticated risk model possible, or build one sophisticated enough to improve decisions while remaining cost-effective to operate at scale.

We chose the latter. Not because we lacked ambition, but because enterprises need to project ROI with confidence. A model that's 5% more accurate but costs 300% more to run doesn't pencil out when you're processing millions of transactions.

The efficiency-first approach enabled us to deploy across consumer lending operations because finance teams could model the business case. The "most capable" approach would have stayed in pilot purgatory, impressive in demos but too expensive to scale.

2. Speed to Value Beats Speed to Market

Enterprises don't adopt AI for its own sake—they adopt it to solve specific business problems with measurable ROI. The question isn't "how fast can you ship the newest model?" It's "how quickly can we deploy this into production and start capturing value?"

Efficiency-driven AI approaches win here because they're:

• Easier to integrate (lower infrastructure requirements)

• Faster to validate (lower cost of experimentation)

• Simpler to govern (clearer cost/benefit analysis)

• More practical to scale (predictable economics)

I've watched clients pass on technically superior solutions because they couldn't get from pilot to production profitably. The friction isn't capability—it's operational reality. Change management takes time. Procurement processes move slowly. Finance needs ROI models that work at scale, not just in demos.

3. Trust Requires Controllability

Here's what took me years to fully appreciate: enterprises aren't just buying capability—they're buying risk management. Every AI deployment is a bet on reliability, accuracy, and controllability under diverse conditions.

Efficiency-driven approaches often provide better controllability because:

• Smaller, optimized models are more interpretable

• Lower operational costs enable more extensive testing

• Clearer resource constraints force disciplined feature prioritization

• Economic sustainability enables long-term support and refinement

When I articulated the business case for automated lending at KeyBank, I didn't frame it as "faster approvals through automation." I positioned it as "risk enhancement through intelligent systems"—technology improving judgment, not replacing it. That positioning enabled adoption because it acknowledged that controllability and trust mattered as much as speed.

The same principle applies to the "scale at all costs" versus "efficiency-driven" debate in AI. Enterprises aren't asking which approach is theoretically more capable. They're asking which approach they can actually control, trust, and operate sustainably.

What the Market Data Shows

The market is starting to reflect this reality. According to recent Menlo Ventures data, Anthropic—with their explicit focus on efficiency and safety—now holds 32% of the enterprise LLM market by usage, surpassing OpenAI's 25%. More tellingly, Anthropic captures 42% of enterprise coding workloads, more than double OpenAI's share.

This isn't happening because Claude is universally "better" on every benchmark. It's happening because Anthropic's efficiency-driven, safety-conscious approach aligns with how enterprises actually want to adopt AI: with clear cost models, controllable behavior, and sustainable economics.

Anthropic generates roughly 8x more revenue per user than OpenAI despite having a fraction of the consumer mindshare. Why? Because enterprise buyers will pay premium prices for AI they can trust, afford to scale, and operate reliably—even if it's not always the "most capable" on paper.

The Implications for AI Strategy

If you're building or buying enterprise AI, the "do more with less" principle has practical implications:

For AI Companies: Stop competing solely on capability benchmarks. Enterprises care more about:

• Transparent, predictable costs

• Reliable performance across diverse use cases

• Clear governance and controllability

• Sustainable economics at scale

• Demonstrated ROI in production (not just demos)

For Enterprise Buyers: Resist the temptation to chase "most advanced." Instead ask:

• Can we model the business case with confidence?

• What's the path from pilot to production at scale?

• How much control do we maintain over behavior and costs?

• What's the total cost of ownership, not just licensing?

• Is this sustainable for our organization long-term?

For Both: Recognize that efficiency and capability aren't opposites—they're multipliers. The most valuable AI solutions deliver meaningful capability improvements while maintaining operational efficiency. That's not compromise; that's strategic optimization.

Why This Matters Now

We're at an inflection point in enterprise AI adoption. The "move fast and break things" approach that worked in consumer markets doesn't translate to enterprises with compliance requirements, risk management frameworks, and P&L accountability.

The companies winning in enterprise AI aren't those spending the most on compute. They're those delivering maximum capability per dollar of investment—which requires algorithmic innovation, not just scaling laws.

This is why Amodei's articulation of "capability per dollar of compute" resonated with me. It's not just Anthropic's positioning—it's the underlying reality of how enterprises actually adopt transformative technology. I've seen this pattern across industries for nearly four decades.

Technical superiority matters. But sustainable efficiency, cost predictability, and operational controllability often matter more. The winners won't be those who build the biggest models. They'll be those who build the right models—powerful enough to solve real problems, efficient enough to deploy at scale, and trustworthy enough to bet businesses on.

That's not settling for less. That's understanding what actually drives enterprise value.

And it's why "do more with less" isn't just a positioning statement—it's the principle that separates pilots from production, demos from deployment, and hype from sustainable business results.