For technology leaders, the key question is not whether SaaS disappears, but who owns the transition from application buying to capability orchestration. If agents sit across ERP, CRM, workflow and analytics, the old procurement model breaks down: vendor choice alone no longer determines value. CIOs need explicit decision rights for where agents can act, what data they may use, and which processes remain human-approved.
This changes governance from system selection to control design. The management issue is whether the organisation has a repeatable way to assess agentic use cases for business impact, operational risk and regulatory exposure. Without that, teams will chase pilot activity while core processes become harder to audit, support and explain. A practical next question is: which workflows are suitable for partial automation now, and which require stronger controls before scale?
Portfolio discipline becomes more important, not less. Leaders should avoid funding isolated AI features inside every SaaS contract. Instead, compare investments that improve data quality, integration and process standardisation against speculative agent pilots. The trade-off is clear: short-term enthusiasm can crowd out the unglamorous work that determines whether AI produces measurable outcomes. Success metrics should include cycle time, exception rates, adoption, and the proportion of work moved from manual intervention to governed automation.
The operating model also needs attention. Moving toward โsystems of actionโ means changing roles for business process owners, data stewards, security, and vendor managers. The winners will be organisations that treat agentic AI as a cross-functional operating capability, not a feature upgrade. The next question for executives is whether their current governance forums can make fast enough decisions without weakening accountability.
Is SaaS really dead?
Some commentators have questioned whether SaaS applications are effectively “dead,” contributing to market volatility for companies such as ServiceNow, Salesforce and Oracle. For example, Alexander Puutio wrote in Forbes, “SaaS Is Dead. Long Live Service-As-A-Service.” Other industry voices have taken similar positions. The narrative has gained traction amid recent earning reports and company guidance that have prompted investors to revise growth expectations for large SaaS providers. The reality is more nuanced. SaaS — and even on-premises applications — are not necessarily going away. What is changing is how they are used.From systems of record to ‘systems of action’
Increasingly, these platforms will function as systems of record, providing the data foundation that drives value from agentic AI. That data still needs to be created, stored and analyzed somewhere — and in most cases, that “somewhere” is still SaaS. What will change in the emerging agent economy is how value will be created and delivered. Instead of simply capturing journal entries or customer records, the agent economy will focus on driving action. According to Peter Ballis, CTO of Workday, agents will help turn ERP from systems of record into systems of action. This shift is already happening. In our research, 67.5% of software companies report that they have already implemented agentic AI solutions. According to Howard Dresner of Dresner Advisory Services, “We do not see agentic AI as eliminating applications or business intelligence systems. Instead, we see it as democratizing access and driving business transformation, as agents reshape what work is performed by humans and what work is automated.” SaaS applications will continue to deliver strong value, but that value proposition will be reshaped by agentic AI. Over time, agentic capabilities will help turn traditional systems of record into measurable systems of value.Guidance for CIOs and investors
The notion that SaaS is dead is simply wrong. According to Google’s survey “The ROI of AI 2025,” 12.5% of 3,466 organizations surveyed reported being early adopters of agentic AI and that they are already achieving measurable value. Our research points to a sobering reality: Only 32% of firms have been successful with business intelligence, and those organizations succeeded by doing the unglamorous work of industrializing data — improving data quality, governance, integration and scalability. Without that foundation, excitement around new technology will not translate into real results. Contrary to some market narratives, this wave of technology change is not about ripping and replacing systems of record. So what should the Street and CIOs actually be concerned about? This wave is erasing boundaries between software categories. For years, there were clear separations among low-code platforms, process development tools, business intelligence, data warehousing and enterprise applications. As those boundaries erode and capabilities overlap, software vendors will find themselves competing with a broader group of vendors. We believe this convergence will drive industry consolidation. For CIOs, the priority is making the right strategic bets. If your organization is still data-immature, focus on tools and approaches that accelerate data maturity and help you move into the agent economy. This is not about buying the flashiest new technology. It is about building communities of practice and developing an overarching strategy that balances tactical execution with long-term strategic change. For investors, the message is similar: Do the homework. Ask which companies are best positioned to enable agentic solutions. Ask who can best help the 68% of organizations that still struggle with data maturity move forward the fastest. Those companies will be the winners — just as Nvidia has been on the chip side. And remember, this shift is not about rip-and-replace. It is about strengthening the foundation needed to move toward true systems of action.Enjoyed this article? Sign up for our newsletter to receive regular insights and stay connected.