AI to ROI Big Story: Why Measuring ROI from AI Is So Hard

2026-01-20 · Source: AI to ROI - By Ray Rike and Peter Buchanan · Field: Business & Management — Corporate Strategy & Leadership, Operations & Process Management, Project & Product Management · Depth: Intermediate, long

Summary

Enterprises are struggling to measure the Return on Investment (ROI) from Artificial Intelligence, despite rapidly increasing expenditures. Ramp data indicates a 13x increase in average monthly AI token spend since January 2025, yet only 27% of executives report meeting their AI ROI expectations. This "AI Measurement Paradox" stems from AI generating "dark output"—meaningful economic value that corporate dashboards fail to capture, such as a legal document cost dropping from \$400 to \$5, appearing only as a token expense. Compounding forces driving costs include a shift to usage-based pricing by vendors like GitHub Copilot and Anthropic, agentic AI devouring tokens (e.g., Uber's 5,000 engineers exhausted their annual budget by April 2026 after agentic usage surged from 32% to 84%), and "tokenmaxxing" where employees maximize AI use without clear goals, with only 18% of coding token spend translating to shipped products. NVIDIA CEO Jensen Huang suggests AI token spend could reach \$250,000 annually for every \$500,000 in developer salaries, making it a compensation-scale expense.

Key takeaway

For Directors of AI/ML or executives evaluating AI investments, your focus must shift from adoption metrics to measurable business outcomes. You should define productivity and set clear goals before deploying AI initiatives, then redesign processes and decision loops rather than simply layering AI onto existing workflows. Implement CFO-level governance for AI spend, treating it as a compensation-scale operating expense. This approach will help you move beyond "AI Dark Output" and realize tangible ROI, avoiding the trap of escalating costs without proportional value.

Key insights

Enterprises struggle with AI ROI due to unmeasured "dark output" and exploding usage-based costs, requiring new measurement frameworks.

Principles

• Measure outcomes, not activity.
• CFO must own AI spend.
• Redesign processes for AI.

Method

An "AI to ROI framework" progresses from cost visibility to usage, then proficiency, and finally business impact, requiring defined productivity, goal setting, and workload-specific measurement.

In practice

• Define productivity before deployment.
• Set clear goals for AI tools.
• Integrate AI cost management tools.

Topics

• AI ROI Measurement
• AI Cost Management
• Agentic AI
• Organizational Redesign
• Productivity Metrics
• Token Consumption

Best for: CTO, VP of Engineering/Data, AI Product Manager, Executive, Director of AI/ML, Consultant

Related on AIssential

• Just What You Need to Determine AI ROI: The Big Book of AI Metrics — Establishing and continuously measuring AI metrics from implementation through operations is crucial for high ROI.
• Exploring AI’s Value and Profitability — AI ROI is about augmenting human work and fostering innovation, not just incremental productivity.
• Don’t Just Talk About AI. Measure Business Outputs. Here’s How. — Measure AI success by business outputs, not activities, to ensure tangible value and avoid pilot failures.
• AI to ROI Metric of the Week: Cost Reduction — AI-driven cost reduction directly impacts profitability and competitiveness, freeing resources for strategic growth.
• If You're Not Seeing AI ROI, Blame Yourself — AI ROI failure stems from a missing board-level strategy, not technology issues, requiring a CEO-led business model re-evaluation.
• The AI Reality Check: What Big Tech Learned the Hard Way in 2026 — Scaling AI from individual productivity to organizational ROI requires robust production systems, data quality, and governance, not just advanced models.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI to ROI - By Ray Rike and Peter Buchanan.