In the digital world,

Are your company’s actual AI costs far exceeding your budgeted costs?

According to a recent survey from SaaS benchmarking firm Benchmarkit and cost governance platform Mavvrik, 85% of organizations misestimate their AI cost by more than 10% and nearly a quarter miss by more than 50%. Many of these cost overruns come from hidden layers beneath the cost of the AI model including data preparation, security, integration, and governance.

The survey also documented that 8 out of 10 companies reported that AI costs eroded gross margins by more than 6% and a quarter of companies saw margins drop by 16% or more.

AI tech stack churn is becoming a cost overrun problem as companies scramble to rebuild their AI infrastructures to keep up with the ever-increasing releases of new product offerings. A survey from AI data quality company Cleanlab found that 70% of regulated enterprises and 41% of unregulated organizations replace at least part of their AI stacks every three months.

In addition, Gartner has documented that SaaS subscription costs from many large vendors have risen 10 to 20 percent over that last year outpacing IT budget growth projections by 2.8 percent. Many vendors are also moving to hybrid pricing models that combine subscription pricing with use or outcome-based pricing.

All these fast-moving changes have made accurate budget forecasts nearly impossible. This new market reality also necessitates a major transition in how companies think about allocating their capital and resources to fund AI going forward.

The need to treat AI like any major capital investment

Treating AI, GenAI, & Agentic AI as major capital investments rather than routine IT expenses is fast becoming a critical measure of success for companies of all sizes, across all industries. Global corporate spending on AI infrastructure and integration is expected to exceed $500 billion in 2026. As all these forms of AI migrate from isolated tools to enterprise-wide integrated systems, companies must manage them as asset-backed resources that require strategic allocation, rigorous governance, and a sustainable ROI. Simply put, C-Suite leaders & Boards of Directors need to stop looking at AI as an IT project and start seeing it as a critical part of their capital allocation strategy and decision-making process.

Here are five principles that forward-looking companies are using to treat AI as a capital investment:

• Shift from a project approach to a portfolio approach: Stop treating AI as a series of isolated experiments and begin managing it as a strategic portfolio of company-wide investments.
• Deploy active allocation: AI does not hold inherent value when it’s not in use; it requires active deployment (like capital) into high-leverage areas, such as workflow redesign, data aggregation & access, new product innovation, and governance.
• Appoint an AI Steward: Firms need a dedicated senior leader to oversee AI capital allocation, performance, and ROI.
• Focus on complimentary assets: AI is not just the model or tool itself. It is leveraged by working with other core enterprise assets such as quality data, robust infrastructure, and workforce training.
• Manage the “J-Curve” of adoption: Like other traditional capital investments, AI often requires substantial upfront investments that may temporally increase costs before they deliver significant ROI.

Managing AI operational costs and governance risk as a single measurable system

AI, GenAI, & Agentic AI are not only expensive to run, they are also expensive to govern. The problem many companies face, however, is the people responsible for operational costs and the people responsible for risk operate in separate organizational silos and often have conflicting priorities.

FinOps is an operational framework that merges finance, technology, and business leaders into one collaborative team with shared accountability for strategic capital allocation, rigorous governance practices, and sustainable ROI results. Working together, these cross-functional teams enable faster product delivery, better financial control & predictability, and ultimately better executive decision making.

Ultimately, FinOps is about getting the most value out of a portfolio of digital technology investments that drive growth & ROI.

Early companies who have adopted FinOps have embraced these 4 key aspects:

• Core Principles: Visibility (knowing what is spent), Optimization (right-sizing resources), and Accountability (teams owning outcomes).
• The Iterative Cycle: FinOps operates in three continuous phases – Inform (strategically allocating investment costs), Optimize (identifying operational redundancies & waste), Operate (Governance & risk protocols and processes).
• Organizational & Cultural Shift: Breaks down silos between finance, technology, and operations to create cross-functional, team accountability rather than centralized, IT accountability.
• Business Value: The goal is not just to reduce costs but to ensure that major technology investments directly drive sustainable growth and ROI.

The 4 Zones Model: A strategic framework for maximizing your portfolio of AI capital investments

To excel at valuation creation as a digital enterprise requires a company to undertake some fundamental changes in its operating model on how it strategically allocates its scare resources and budgets. In working with forward-thinking leaders to help them with this shift, I have relied on my brother, Geoffrey Moore’s, 4 Zone model framework shown above.

We have used this framework to identify and prioritize multiple business-value creation needs and opportunities across different operating and functional units. We utilized inputs from cross- functional teams to rigorously explore how deploying AI, GenAI, or Agentic AI could deliver the desired business outcome and ROI.

Productivity Zone AI investments focused on eliminating bottlenecks and redundancies in critical operating systems and streamlining organizational workflows.

Performance Zone AI investments focused on enhancing employee and customer experiences and aggregating and leveraging centralized data resources.

Incubation Zone AI investments focused on expediting speed to market and time to value for new products and services.

Transformation Zone AI investments focused on net new lines of business that could scale to material revenues and profits.

Companies continue to wrestle with how to successfully navigate the challenges and opportunities presented by the unprecedented new wave of AI technologies. As such, it is becoming clear that traditional operating models, resource & budget allocation processes, and legacy systems & workflows must undergo fundamental changes. A good place to start is to treat AI in all its forms strategically, not tactically.