H1 2026 earnings season told the story vendors wanted told: AI attach rates climbing, cloud backlog compounding, and platforms expanding into adjacent workflows.

ServiceNow sustained approximately 19% subscription growth while increasing its 2026 AI commitment to $1.5 billion. Salesforce grew Agentforce annual recurring revenue to $1.2 billion, even as core revenue growth slowed into the single digits. SAP reported approximately 30% Cloud ERP Suite growth in constant currency. While numbers are impressive, they don’t demonstrate customer value and they don’t tell the full story.

Underneath the growth narrative, enterprise software vendors are rewriting the terms of buyer leverage, and they are doing it while customers deploy, not while they negotiate. In fact, there are three important shifts in buyer leverage that CIOs need to prepare for now before the next renewal.

  • Today’s consumption pricing is becoming the new floor for software renewals. Credits, units, and usage tiers give buyers flexibility during deployment, they also give vendors a baseline at renewal. Consumption pricing lowers barriers to adoption, but over time usage patterns become reference points for budgeting, forecasting, and future negotiations. Every ungoverned pilot can become next year’s committed floor. It’s critical for CIOs to secure spending controls, rollover rights, metering transparency, and unit-price protections while usage remains small enough to negotiate.
  • Control of AI value measurement is becoming another source of leverage. Vendor-reported AI metrics are designed to demonstrate momentum to investors and of course, to shape the overall narrative. They want buyers to accept the narrative. But metrics about adoption or activity say little about effort removed, cost avoided, quality improved, or business outcomes achieved. Measure AI against operational results, not adoption statistics, or you risk funding someone else’s success story. Don’t focus on how many credits were consumed or agents were activated, focus on whether AI is helping you complete work faster, better, or at lower cost.
  • Workflow dependency is becoming the new lock-in. Switching costs are moving out of databases and into orchestration logic, semantic models, configurations, permissions, and agent behavior. Your data may be portable. Your workflows increasingly are not. As AI becomes embedded inside business processes, portability becomes an operating concern, not simply a technical one. CIOs should ensure that workflow definitions, orchestration artifacts, semantic mappings, and agent histories remain as portable as the underlying data.

Make Vendor Dependency An Executive Decision

Of all the shifts in buyer leverage, workflow dependency is the one CIOs should be the most concerned about because it impacts overall strategy, platform choices, and architectural choices. Not all dependencies are bad. Standardized platforms can accelerate deployment, simplify governance, and reduce integration complexity. The mistake is accepting dependency without understanding both its value and its exit cost.

Over the next 90 days, identify the five AI-enabled workflows most likely to become operationally essential. Calculate the value they create, the consumption exposure they generate, and the cost of reconstructing them elsewhere.

Then classify every dependency as: a.) Intentional, b.) Negotiable or c.) Unacceptable. Scale the dependencies whose value justifies their cost. Reset the pricing, usage, or exit terms on those worth keeping. Stop funding the rest.

From here, CIOs need to work with the entire ecosystem. CIOs and their technology leaders must own consumption telemetry (e.g., credits consumed, API calls, tokens used, agent invocations etc.) while partnering with both Sourcing leaders who must own portability and renewal protections in contract negotiations and Finance leaders who can enforce spending ceilings.

Some level of vendor lock-in is going to be unavoidable, but every concession of flexibility must come with measurable business advantage and, more importantly, customer value.

AI is changing the terms of the deal between buyers and vendors. It is determining who captures the economic value created by AI. That power compounds quietly, quarter by quarter, inside decisions that are rarely framed as negotiations. It’s on CIOs to be sure they understand exactly which dependencies they are creating, what those dependencies are worth, and what it would cost to walk away.

Forrester clients should read our detailed report and schedule a guidance session. We will help identify where leverage has already shifted, where it remains recoverable, and where governance actions today can prevent commercial constraints tomorrow.

Thanks to the contributions and expertise of our team that helped shape the analysis, insights, and recommendations presented in this report — Stephanie Balaouras, Kate Leggett, Leslie Joseph, Joe Cicman, Akshara Naik Lopez, George Lawrie, and Hannah Murphy.

Share