If you listened to the enterprise software earnings calls, you’ve heard them flooded with AI growth claims — “AI‑influenced revenue,” expanding cloud backlogs, and agentic workflows. From the vendors perspective, these confident assertions sound like were in some golden age of digital transformation. But let’s be clear, an organization’s reality is messy and more complex. Enterprise AI adoption tells a different story. Deployment is still early, uneven, and often subsidized by organizations through pilots, services, and internal churn. The real story, and where things get interesting, are these billing models shift the financial risks to enterprise budgets, and at the same time, vendor claims can turn into vulnerabilities and procurement leverage.

The Adoption Gap Hidden Inside AI Earnings

The uncomfortable truth is that a large share of reported AI growth blends genuine innovation with reclassification. Vendors increasingly bundle rebranded data clouds, existing tools, and legacy SKUs into “AI‑influenced revenue” a pattern we detail in in our H2 2025 earnings report.

In practice, most enterprise “success” still looks like controlled experimentation. Key takeaways:

  • Pilot-heavy, production-light deployment. Accuracy and productivity claims remain tied to supervised pilots rather than uncontrolled global environments.
  • Deal sizes that signal testing, not transformation. Deals clustering around $100,000 to $200,000 almost always indicate validation cycles, not enterprise rollout.
  • Bundling that hides real consumption. “AI-influenced revenue” often includes old SKUs repackaged as AI, obscuring what was actually used and what business value was realized.

This gap used to be survivable. It isn’t anymore because vendors are more likely attaching AI to metered pricing models that push volatility directly onto your budget.

Govern Usage-Based AI Before It Becomes A Silent Tax

The biggest structural shift in H2 2025 earnings is the move from seat‑based licensing to consumption‑based AI monetization credits, tokens, and metered usage. Vendors benefit because these models stabilize revenue, but it forces you to absorb the volatility as spend becomes sensitive to data patterns, workflow shifts, and opaque metering triggers.  If you aren’t careful enterprise software AI spend will start behaving like runaway cloud infrastructure costs. Best practice is to control and govern AI consumption the same way you govern cloud infrastructure, a practice we outline in our guidance on Cloud FinOps.

  • Put a ceiling on exposure. Negotiate annual caps and explicit coverage rules to prevent runaway usage from becoming unplanned operating spend.
  • Stop “use-it-or-lose-it” inflation. Push for rollover protection. Otherwise, unused credits silently raise your next renewal baseline.
  • Demand real-time visibility. Require dashboards that map burn rates to specific use cases and business owners, not aggregated consumption blobs.
  • Expect tightening pricing mechanics. Vendors under growth pressure will shorten expiration windows and refine metering. Build scenarios where token burn rises by double digits even without new use cases.
  • Use vendor pressure as a commercial tool. When growth slows or backlog outpaces revenue, vendors become much more flexible. We cover these signals in depth in our Agentic Business Fabric report.

Use Earnings Pressure As Procurement Leverage

Earnings season is more than sharing business performance, it shows where vendors hold power and exposes where they’re stretched thin. Vendors growing below ~15% face investor urgency and that tension is your negotiating leverage.

Backlog patterns matter even more. When contracted backlog grows faster than revenue, sales are outrunning delivery. That imbalance creates implementation risk and negotiation opportunity.

Your moves heading into renewals:

  • Segment vendors into leaders and laggards. Laggards should pay more for your logo, references and AI case studies.
  • Exploit backlog-revenue gaps. Secure delivery SLAs, named resources, and penalties tied to implementation risk.
  • Use on-premises footprint as leverage. Especially when vendors push cloud-only positions. Your installed base remains a bargaining chip, not a liability.

Redefine The Narrative To Your Advantage

AI agents, consumption pricing, and consolidated data layers all shift risk away from vendors and toward enterprises. Treat every AI claim as unproven until validated in your internal environment. Govern consumption with financial discipline. Contract for optionality before data gravity locks you in. Don’t wait for your next renewal cycle to start asking these questions. The pricing models solidifying in these earnings calls will shape your cost structure for the next few years.

Forrester clients can schedule a guidance with me to pressure‑test your enterprise software investment assumptions, SaaS governance and contract renewal strategy.

Read the full report: Enterprise Software Latest Earnings In H2 2025 Is The AI Hype Versus Procurement Power Battle