Ben Pippenger, cofounder and Chief Strategy Officer at Zylo.
Earlier this year, when Microsoft CEO Satya Nadella suggested that software-as-a-service (SaaS) applications may “collapse” in the AI agent era, the tech world understandably did a double take. On the surface, it’s a provocative thesis: If AI can do the work, why do we need apps?
But look a little deeper, and the picture becomes more complex.
AI isn’t going to kill SaaS. But it is going to make SaaS a lot more chaotic, fundamentally changing how it’s bought, used and measured.
SaaS Isn’t Dying—It’s Transforming With AI
Enterprise business-to-business (B2B) SaaS isn’t disappearing—it’s evolving. Leading platforms are already weaving AI into their core experience, whether through embedded companion assistants or integrated capabilities.
Meanwhile, demand for SaaS remains strong. Organizations across industries are adopting new tools and spending more on software. The software layer is still critical, but now it serves as the foundation where AI does its work. What’s changing is how that work flows—and how we manage it.
And that’s where things get complicated. AI is fundamentally altering how work gets done. That shift forces a rethink of how we measure cost, value and accountability.
AI Is The New Cost Center—And It’s Harder To Track
For decades, people were the primary cost of getting work done. Headcount was your largest line item—and one of your most predictable. You were paying for time, skill and output.
That equation is changing. Increasingly, AI tools and services are doing the work—not just supporting it, but executing it. Teams aren’t being fully replaced, but the economics of where spend occurs are shifting. You’re still paying for work—but not via salary. Is it per token? Per API call? Per outcome?
Without a transparent way to measure impact, AI can become just another black box—one that silently executes tasks and incurs costs.
The old cost centers—people, infrastructure, software—are being reshaped. AI doesn’t just automate work—it fragments it. Your biggest cost may soon be a collection of AI agents, each handling micro-tasks, each priced differently.
SaaS Pricing Is Fragmenting In The AI Era
As AI becomes more embedded in enterprise software, it’s also reshaping how SaaS is licensed. Traditional seat-based models are giving way to usage-based or hybrid pricing. In theory, paying only for what you use sounds efficient. In reality, it’s harder to control—and even harder to forecast.
Vendors define “usage” in inconsistent ways: one might bill per task, another per resolution and yet another per document or workflow. There’s no standard—and often, no clear way to predict spend.
Usage data is also fragmented. It’s scattered across admin panels—if it’s available at all. Many vendors don’t expose this data through APIs. Even when accessible, usage is often extracted manually and hard to interpret.
Add in fast-changing SKUs and complex billing models, and surprises become inevitable. According to my company’s 2025 SaaS Management Index, more than 66% of IT leaders have been caught off guard by AI- or usage-related charges.
When visibility disappears and pricing becomes opaque, a bigger question emerges: How do you define value?
Redefining Value In A Usage-Based AI Economy
At its core, this shift demands a new definition of value. The old model—$X per user, per month—no longer applies. In today’s AI-powered enterprise, productivity is measured in units of work.
To keep up, organizations need better ways to track return on investment (ROI). That means clearer systems to monitor usage—and tooling that reveals what’s been invisible until now.
The Hard Questions Smart SaaS Leaders Are Asking Now
If you’re a CIO or IT leader, now is the time to ask tough questions:
• Are we paying for outcomes—or just usage?
• Are we tracking value consistency across our stack?
• Are we seeing a unified view of all SaaS and AI tools—who’s using them, how often and at what cost?
• Are we forecasting SaaS and AI spend with confidence?
• Are we catching anomalies before they hit the budget?
• Are we using consumption trends to strengthen renewal negotiations?
• Are we making usage data available and understandable across teams?
• Are we managing AI and SaaS sprawl with clear governance processes?
• Are we assigning clear ownership of usage-based tool oversight?
• Are FinOps, IT, procurement and software asset management (SAM) teams aligned?
Visibility And Governance Will Define The Winners
Leaders who can answer these questions confidently will be the ones who stay ahead. The next era of software won’t be defined by how much AI you implement—it will be defined by how well you measure and govern it. That means knowing not just what your tools do, but what they deliver. Not just what they cost, but what they’re worth.
B2B SaaS is evolving fast. The winners will be the ones who understand what AI is doing, how much it’s costing—and why it matters.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
