AI Survey Program: Cost vs. Value and the Interaction Pattern Shift

As enterprises rapidly scale AI, the market has measured spend extensively, but measured value almost not at all.

This survey explores how technology leaders are navigating AI investments, cost predictability, and true business return. We are testing a core thesis: ROI isn't just about "cost per token," but rather the maturity of the workflows those tokens land in , from basic assisted tools to fully native, agentic interactions.

Your insights will help build a new industry benchmark for AI ROI. All responses are strictly confidential and aggregated.

Thank you for your time and perspective.

Key Definitions for this Survey

(Please review briefly before beginning)

• AI Assistant or Copilot: A conversational AI tool people use to draft, summarize, research, or analyze (either standalone or built into existing software).
• AI Agent: An AI system that can plan and execute multi-step tasks with limited human involvement, completing a process end-to-end rather than answering a single prompt.
• Fixed AI Spend: Per-seat or flat subscription fees paid regardless of how much the tool is used.
• Variable AI Spend: Charges based on usage, such as API calls, tokens, queries, or agent runs.
• Token Governance: The policies and controls an organization uses to manage AI—such as usage caps, approval steps, model routing, and business chargebacks.