Enterprise SaaS is Changing
The Death of Enterprise SaaS?
Every few years, someone declares the death of enterprise SaaS. The claim usually resurfaces when growth slows, multiples compress, or a crack appears in the armor of incumbents. This time, admittedly, the prognosis feels more serious. The rise of reliable internal AI agents and the rapid fall in software development costs are very real. They are structural forces that directly attack the moat of enterprise SaaS as it has existed for the past two decades.
For most of its history, enterprise SaaS solved three problems better than internal IT ever could. It abstracted complexity, delivered continuous upgrades, and spread development costs across thousands of customers. That bundle justified recurring subscription pricing and produced some of the most valuable software companies in history. What is changing now is not demand for software, quite the opposite, but who will be providing it. As that shifts inward (think internal tools, agents, etc.), the usefulness of externally packaged software declines.
The main driver of this shift is the massive decline in software development cost. Modern development tools have dramatically reduced the time and talent required to ship production-grade tools. Tasks that once required entire engineering teams can now be handled by small internal product groups augmented by AI-assisted coding. Enterprises are already building internal applications that replicate meaningful portions of what they previously licensed from vendors.
At the same time, generative AI is changing how software is consumed. Traditional SaaS products are designed around static interfaces and predefined workflows with limited customizability. Did your organization purchase Workday or Salesforce? Time to onboard employees to learn how to use the software itself in its current form.
Internal AI agents invert that model. Instead of users adapting their behavior to the software, agents adapt software behavior to the user. They sit on top of proprietary data, reason across multiple systems, and execute tasks end to end. When an agent can pull data from a CRM, trigger actions in a workflow engine, generate documentation, and summarize outcomes without human intervention, the value of the individual tools beneath it becomes less visible and less differentiated.
It could be argued that SaaS adoption in large enterprises has largely reached saturation. Organizations already run dozens, if not hundreds, of subscriptions. Growth for enterprise SaaS companies comes from upsells, seat expansion, and pricing power rather than net-new category creation. It has and always will be the “land-and-expand” model. It is the reason why growth-stage investors care so much about expansions, net-dollar retention, and other revenue quality metrics.
AI adoption, by contrast, is still in its early innings. Enterprises are investing aggressively in internal models, data platforms, and agent frameworks precisely because these assets compound internally. The result is beginning to look like an inverse adoption curve. Enterprise SaaS adoption is flattening just as internal AI adoption accelerates upward.
What does the market think?
Public markets appear to be recognizing this shift. Even best-in-class SaaS incumbents are seeing meaningful multiple compression, not because they are failing operationally, but because their long-term relevance is being repriced.
Salesforce, the marquee enterprise SaaS success story, still benefits from unparalleled distribution and a deeply entrenched CRM footprint. Yet its growth profile has slowed materially, and the stock is down roughly 18% over the past year. The company’s AI narrative has focused on augmenting existing workflows rather than redefining the economic role of the platform. Investors appear skeptical that layering AI assistants onto a mature product restores long-term pricing power in a world where intelligence increasingly lives inside the enterprise.
ServiceNow is arguably better positioned, given that its core value has always centered on workflow orchestration rather than static software. That alignment matters in an agent-driven future. Even so, its stock is down approximately 31% over the past year. The market seems to be pricing in a harder question. If orchestration logic increasingly resides inside enterprise-built agents that span multiple systems, the advantage of a centralized external workflow layer narrows. ServiceNow’s risk is not outright displacement, but sustained margin and relevance compression.
Workday’s position is more defensible, but only at the core. As a system of record for financials and human capital, it benefits from regulatory complexity and deep process embedding. That insulation is real. Still, the stock is down roughly 17% over the past year, reflecting pressure on the layers around the record. Reporting, analytics, and decision support are precisely the domains where internal agents excel. As intelligence migrates upward, the differentiation of peripheral SaaS modules weakens, even if the underlying data remains sticky.
Atlassian illustrates the most acute exposure. Its products sit at the center of developer and collaboration workflows, the very environments where internal agents are advancing fastest. Developers are among the earliest and most aggressive adopters of internal AI tooling, and collaboration itself is being reshaped by AI-mediated interfaces. Atlassian’s stock is down approximately 40% over the past year, suggesting the market is pricing in a future where its tools are no longer the primary coordination surface, but inputs into internally controlled intelligence layers.
Enterprise SaaS is not going away, but it is no longer where most value is created. In the past, SaaS products defined workflows and controlled how work was done, which gave vendors strong pricing power. As companies build internal AI agents, SaaS increasingly sits in the background, supplying data and basic functionality. The real advantage shifts to whoever controls the intelligence layer on top, making SaaS harder to differentiate and monetize at a premium.