AI Budgets Are Growing Faster Than Business Value

The AI Conversation Is Moving From Adoption to Accountability

For several years, leadership teams were rewarded for moving quickly: give employees access to AI, launch pilots, test agents, add copilots and experiment with new models. That learning phase was useful. It is also becoming insufficient.

The management question is changing from whether the organization is using AI to which AI investments are actually earning the right to scale. That is not primarily a model-selection problem. It is a product portfolio problem.

Adoption Is Rising Faster Than Deep Operating Integration

Statistics Canada reported that 19.2% of businesses used AI to produce goods or deliver services during the previous 12 months, up from 12.2% in 2025 and 6.1% in 2024.

Bank of Canada research published in August found that more than two-thirds of surveyed business leaders personally use AI tools in a typical work week, while only 8% of surveyed businesses reported significant AI use in core operations. A company can therefore have extensive AI activity without having an AI operating advantage.

Spending Is Becoming a Decision Problem

Ramp's August AI Index reports continued U.S. business AI spending while arguing that businesses may be reaching limits on how much additional model performance is worth paying for. Its data shows growing use of cheaper models and model-serving platforms as adoption growth among major providers slows.

IBM's August launch of Apptio AI Value & ROI is another market signal: technology and finance leaders increasingly need to connect AI costs with outcomes across revenue, cost, speed, productivity and risk. Cost visibility alone is no longer enough.

What Organizations Often Get Wrong

AI investment gets distorted when leadership treats every initiative as part of one undifferentiated budget, measures activity instead of value, lets pilots become permanent by inertia, and evaluates each experiment without comparing its opportunity cost against other product and operating investments.

• Seat adoption and token usage can explain activity but do not prove value.
• A demo proves technical possibility, not customer adoption or economic value.
• A pilot without explicit evidence thresholds can continue indefinitely.
• Every team and dollar allocated to one initiative reduces capacity for another opportunity.

Manage AI as a Product Portfolio

For every material AI investment, leadership should use a common decision system.

• Define the consequential customer or operating problem.
• Assign a business or product owner accountable for the outcome.
• Define the measurable result and establish a baseline before implementation.
• Calculate total economics beyond licences or model charges, including data, integration, evaluation, security, support and change management.
• Separate assumptions from observed evidence.
• Identify risks that become material only at scale.
• Make an explicit Scale, Validate Further, Redesign or Stop decision.

The AI Portfolio Review

Leadership teams do not need another steering committee that only reviews project status. They need an investment review that compares initiatives using common dimensions such as problem value, strategic alignment, measurable outcomes, evidence strength, data readiness, feasibility, adoption probability, economics, operational ownership and risk.

The purpose is not to create a mathematically perfect ranking. It is to make tradeoffs visible and prevent immature experiments from receiving the same funding logic as initiatives with strong evidence and clear economics.

The Product Strategy Connection

Good product strategy already asks the questions AI programs now need: which problems matter most, which users are being served, which outcomes are worth pursuing, what assumptions are being made, what evidence supports them, where scarce capacity should be allocated, and what the organization will deliberately not do.

AI does not remove the need for those decisions. It makes the cost of avoiding them larger because experimentation can proliferate so quickly.

Where PeterPaps Can Help

PeterPaps helps SaaS and B2B technology organizations turn ambiguous opportunities into explicit product and investment decisions.

A Product Strategy engagement can help inventory AI initiatives, define comparable decision criteria, connect investments to customer and business outcomes, identify evidence gaps and build a roadmap around the opportunities most likely to create value. Fractional Product Leadership can provide the operating discipline to keep those decisions connected to execution as evidence changes.

Related Insights

• Your Team Uses AI. Your Business May Not Be
• AI Spending Is Becoming a Product Portfolio Problem
• Your Roadmap Is Busy. That Does Not Mean It Is Strategic