Skip to content
Fast Company Executive Board

The Fast Company Executive Board is a private, fee-based network of influential leaders, experts, executives, and entrepreneurs who share their insights with our audience.

The next AI maturity marker

From adoption to optimization, the next phase will drive innovation without compromising efficiency and sustainability.

[Photo: Getty Images]

Profile photo of Aviad Almagor

Vice President of Technology and Innovation · Trimble

The first wave of artificial intelligence adoption was driven by excitement. Organizations rushed to deploy generative AI tools, automate processes, and explore what could be done faster with AI. Leadership teams encouraged experimentation and curiosity-fueled adoption; no one wanted to be late to the game.

Now, as adoption scales, companies are facing a new reality—one where the cost of compute, environmental impact, and business model alignment matter as much as innovation.

The next chapter is less about how fast organizations adopt AI and more about how intelligently they optimize it. Internally, that means building strategies that extract maximum value from AI compute—in other words, efficiently managing the cost of AI. Externally, it means ensuring AI-powered offerings drive meaningful, sustainable revenue. Both sides of this coin are part of the evolution toward smarter, more sustainable AI use.

THE AI COST PARADOX

Historically, SaaS growth followed a simple logic: More engagement meant more value. But, with generative AI, usage comes with real compute costs, and those costs scale fast. According to Gartner, organizations that incorrectly model usage could miscalculate AI costs by 500% to 1,000%. Every prompt, token, or model carries a price. A single developer using an AI coding assistant can incur thousands in compute charges in just weeks. Multiply that across departments, and enthusiasm can turn into a runaway budget.

The irony: The most active AI users can become the most expensive ones.

OPTIMIZE AI INTERNALLY

Organizations that jumped headfirst into generative AI are now confronted with the real-world costs of compute usage, model fine-tuning, scaling infrastructure, and the training required to help teams use AI effectively. 

Internally, optimization begins with education and visibility into usage and spending. Employees should understand that every token processed has a cost. For example, employees could have access to a dashboard that displays the accumulated cost of their work. Using the right model for the task at hand is another way to optimize AI usage; not every task requires the most advanced model available.

Leaders should prioritize use cases where AI drives meaningful outcomes versus novelty. If AI doesn’t move the business forward in some way, it doesn’t belong. AI should be used where it can deliver faster results, automate workflows, or provide insights that humans can’t. Gartner refers to this as strategic IT cost optimization—not just cutting costs, but shifting spend toward high-impact initiatives and continuously monitoring value.

In addition to lowering operational costs and accelerating return on investment, optimizing AI has a positive impact on sustainability. Every prompt requires energy. Every model training cycle consumes massive compute resources. Compute costs aren’t just financial; they are also environmental.

Even small efficiencies make a difference. More concise prompts, fewer unnecessary iterations, and smarter model selection can help minimize AI’s environmental footprint.

RETHINK AI BUSINESS MODELS

As companies optimize AI internally, they must also rethink how they offer it to customers. Bundling AI into standard subscriptions or charging a fixed premium fee becomes challenging as compute costs naturally rise with greater usage. Today, several pricing models are emerging and aim to offer flexibility without sacrificing predictability:

Usage-based: Organizations pay for AI based on actual consumption, such as tokens, which ensures that costs scale in direct proportion to usage.

Tiered: Volume-based discounts provide predictability and enable customers to scale their AI usage effectively. 

Value-based: Ties costs to measurable outcomes, such as productivity gains or time saved, to align pricing with the real business value derived from AI.

Hybrid: Blends subscription fees with consumption-based charges to help organizations budget, offering the option to pay more only if usage increases.

The healthiest models align value creation with cost generation for both the vendor and the customer. In other words, customers pay more when they receive more, and vendors earn more when they deliver real results. This alignment ensures that both sides benefit as AI adoption continues to grow.

SMART ADOPTION, SMARTER OPTIMIZATION

The next phase of AI adoption is about striking a balance—driving innovation without compromising efficiency and sustainability. Success metrics are shifting from raw usage to real value, focusing on cost per insight, compute per outcome, and environmental impact. AI’s promise will be fully realized not by how widely it’s adopted, but by how intelligently it’s optimized, both within organizations and across the broader business landscape.

Aviad Almagor is vice president of technology and innovation at Trimble.