Walk into any investor meeting at SAP, Oracle, Salesforce, or Workday this quarter, and you’ll hear the same story. The enterprise software vendors have discovered their future: they’re no longer mere software companies, but strategic partners in your AI transformation. They’ve launched orchestrated agent architectures, embedded foundation models, deployment frameworks, and—most importantly—consulting services bundled into their offerings. This pivot, they’ll tell you, represents visionary leadership and strategic foresight. The market has a different interpretation: desperation masquerading as reinvention.
The enterprise software vendor industry faces a structural crisis that the industry’s own narrative carefully obscures. Their traditional licensing model—charging per user for software that increasingly gets replaced by AI agents—is becoming economically obsolete. Rather than acknowledge this market-driven shift, vendors are cycling through pricing models, intensifying architectural lock-in, and bundling consulting services while simultaneously losing confidence in their core products’ ability to deliver measurable value.
The data tells a clearer story than the press releases. Across 40 major SaaS earnings calls in 2025-2026, executives admitted that AI revenue contributions remain “margin neutral” at best, while “infrastructure costs are just very expensive at the moment.” Translation: they’re selling AI services without knowing how to make them profitable. Meanwhile, 95% of enterprise AI pilots fail to scale, and only 5% deliver measurable profit impact. When 43% of organisations cite uncertainty measuring return on investment and 55% struggle with AI reliability and hallucination management, the problem isn’t adoption—it’s that the software doesn’t work reliably enough to replace existing decision-making processes at scale.
Yet vendors continue jacking up prices. Microsoft implemented 25% to 33% price increases on frontline worker licenses in July 2026 while removing automatic volume discounts. Adobe raised Creative Cloud pricing by up to 27%. VMware, now facing customer revolt, attempted price increases ranging from 150% to 1,200% as perpetual licenses convert to subscription-only. These aren’t strategic price adjustments in a thriving market. They are the actions of vendors watching their traditional revenue model collapse and overcompensating through aggressive pricing before customers leave entirely.
The pivot to “AI consulting” is the visible symptom of this deeper crisis. Salesforce alone cycled through three entirely different Agentforce pricing models in less than 18 months. This isn’t sophisticated pricing architecture—it’s vendor confusion and customer rejection cycling in real time. As one Salesforce executive candidly noted, “any vendor who thinks they have it all figured out is kidding themselves.” Translation: we’re guessing at pricing because we have no idea what customers will actually pay for AI agents that haven’t proven they deliver value.
The consumption-based pricing shift exposes the problem most clearly. For decades, enterprise software vendors charged per user, anchoring revenue to headcount. This model assumed predictability: more employees meant more software seats meant stable, recurring revenue. AI fundamentally breaks that assumption. If ten AI agents can perform the work of one hundred sales development representatives, why should a company maintain one hundred CRM licenses? The answer is it shouldn’t. Vendors know this. So they’ve abandoned per-user pricing in favor of consumption-based models charging per transaction, per agent action, or per conversation.
This switch reveals vendor capitulation, not strength. Consumption-based pricing essentially says: “We no longer believe customers will pay predictable fees because we can’t guarantee consistent value delivery.” If a vendor had confidence their software was delivering measurable return on investment, they’d maintain per-user pricing and guarantee renewal. Instead, they’re shifting variable cost risk to customers and tying revenue to usage—a classic hedge against customer churn and a signal that traditional licensing power is eroding faster than the vendor rhetoric suggests.
Customer behaviour confirms the crisis. When enterprises have a choice between corporate-mandated tools and alternatives, the numbers are brutal. Adoption drops from 68% when tools are forced to just 18% when alternatives are available. Microsoft’s Copilot loses to ChatGPT 76% to 24% when employees can choose. This happens despite Microsoft’s enormous distribution advantage and aggressive bundling. Employees, it turns out, don’t trust vendor-built AI agents when other options exist. This is not a sales and marketing problem. It’s a trust and capability problem that no amount of consulting services bundling will solve.
The enterprise response has been equally telling. Organisations facing AI adoption challenges are reversing the traditional “build versus buy” calculus. As AI infrastructure commoditises and deployment costs fall, building internal AI platforms becomes financially superior to renting from vendors over a five-year horizon. Enterprises gain ownership, protect margins, avoid vendor lock-in, and build proprietary capabilities rather than depending on vendors’ next pricing cycle. Some have begun quietly migrating away from legacy platforms or slowing renewal cycles, waiting for vendors to stabilise pricing and prove value rather than capitulating to aggressive increases.
India’s technology services sector provides a canary-in-the-coal-mine signal of how structural this shift really is. Infosys cut 25,994 employees in 2024—its first annual headcount decline since 2001. TCS reduced 13,249 employees, Wipro shed 21,800, and collectively India’s major IT firms eliminated over 50,000 jobs. These are not cyclical adjustments. They represent the collapse of the traditional effort-based staffing and billing model that underpinned decades of IT services growth. HCL, which focused on AI-led delivery transformation rather than volume staffing, was the lone major firm that added employees. The broader Indian IT sector faces 2% to 3% annual revenue deflation as AI-driven productivity compression makes traditional billing models economically unviable. For vendors who historically relied on Indian system integrators for implementation, this creates a cascading crisis: their own implementation partners are shrinking, forcing them to either build internal consulting capacity or watch implementation margins compress further.
Vendors’ response to this crisis has been to intensify lock-in through proprietary orchestration frameworks. AWS’s AgentCore, Salesforce’s Koa, and similar vendor-specific agent runtimes are far more proprietary than any legacy SaaS ever was. Once organisations build workflows on these frameworks, switching becomes exponentially harder than migrating from traditional software. The data accumulates, the agents get fine-tuned on proprietary data, the integrations deepen. Combined with rising prices and high switching costs—58% of organisations attempting AI platform migrations experience problems—vendors are creating financial entrapment as a substitute for product confidence. This is not a sign of vendor strength. It’s a sign that vendors have stopped believing they can compete on merit and are instead betting they can lock customers in before the alternatives become obvious.
The headline narrative about “Enterprise software vendors pivoting to AI consulting leadership” masks a strategic admission of defeat. Accenture, Deloitte, and EY still dominate the $11 billion AI consulting market. Vendors haven’t captured it because implementation at scale requires orchestrating technology change, organisational design, process transformation, and talent development simultaneously. Software vendors are excellent at selling software. They’re discovering they’re not particularly good at running transformation consulting practices. So they’re bundling AI services into existing contracts, raising prices, and hoping customers won’t notice they’re paying for consulting services that will mostly be delivered by the system integrators they’re claiming to partner with.
For CEOs and CFOs, this moment presents unusual opportunity. Enterprise software vendors are transitioning from strength to defensiveness precisely when they’re raising prices and bundling unproven services. Their traditional licensing model is collapsing, their AI projects are failing to deliver value at scale, their own customers are building alternatives, and their implementation partners are shrinking. This is when pricing power typically evaporates and negotiating leverage shifts dramatically to buyers.
The vendors will continue telling you they’re leading the AI transformation. Their actions—cycling through pricing models, intensifying lock-in, bundling services they can’t execute well, hiding behind proprietary frameworks—tell a different story. They’re abandoning confidence in their core business and scrambling for defensive positioning before market forces strip away their pricing power entirely.
The window for negotiating from that leverage before they’ve locked you in through proprietary frameworks is closing faster than the vendor rhetoric suggests.