The modern enterprise software boardroom is currently haunted by a single, terrifying realization: the seat is disappearing. For two decades, the SaaS industry operated on a simple, scalable logic where growth was measured by the number of licenses sold. If a company grew its headcount, the software vendor grew its revenue. But as AI agents transition from simple chatbots to autonomous workers capable of executing complex workflows, the fundamental unit of value in software is shifting from the human user to the digital outcome. This shift is not a theoretical concern for the future; it is already manifesting as a brutal correction in the public markets.

The Great SaaS Repricing

Recent market data reveals a systemic devaluation of the software-as-a-service sector. Public SaaS stocks have experienced a sharp repricing, with the median stock price plummeting by 32 percent. This is not a temporary dip but a fundamental collapse in how investors value software companies. The enterprise value to revenue multiple, a gold standard for SaaS valuation, has shrunk from 9.1x to 4.8x, representing a 42 percent contraction. The scale of this correction is nearly universal, with 86 percent of all tracked stocks experiencing a decline in their multiples.

Interestingly, this crash is not tied to a lack of growth. When analyzing 130 different stocks, the data shows that vertical SaaS—software tailored for specific industries—and horizontal SaaS—general-purpose tools—are growing at almost identical rates. The median growth rate over the past year was 14.1 percent for vertical SaaS and 14.7 percent for horizontal SaaS. Furthermore, the correlation between revenue growth and stock price is a negligible 0.07, while the correlation between EBITDA margins and stock price is actually negative at -0.03. This suggests that traditional financial performance metrics are no longer the primary drivers of market value.

Instead, the market is applying a new set of survival criteria to determine who deserves a premium and who is headed for obsolescence. Investors are now scrutinizing six core dimensions: the presence of a proprietary data flywheel, the alignment of pricing models with AI value, the replaceability of existing workflows, the reliability of the AI implementation, the inherent complexity of the domain, and the strength of the agent ecosystem.

The Death of the Regulatory Moat

For years, many vertical SaaS companies claimed a moat based on regulatory barriers. They argued that because their software handled complex compliance or industry-specific legal requirements, they were protected from competition. The market once rewarded this with a massive premium, but that logic has collapsed. The premium for companies relying solely on regulatory moats has evaporated, crashing from 120 percent to a mere 15 percent. In the age of AI, regulatory knowledge is no longer a secret; it is a dataset that can be ingested and replicated by a sufficiently powerful model.

This has led to the rise of the vertical halo discount. Companies that marketed themselves as industry-specific without possessing deep, exclusive data are now trading at a 40 percent discount compared to their horizontal counterparts. The market has realized that being vertical is not a strategy in itself; it is merely a category. The real value lies in the data that the software captures.

This is where the divide becomes stark. Companies that possess truly proprietary, non-public domain data continue to command a 72 percent premium. This data is the only remaining defense against the commoditization of software. Because AI agents require high-fidelity, expert-level context to make decisions—data that does not exist on the open web and cannot be bought—the platforms that sit on top of these exclusive data streams become the essential infrastructure for the agentic era.

This shift is most visible in the divergence between pricing models. The industry is witnessing a violent pivot away from seat-based licensing toward usage-based billing. Companies like Bandwidth, MongoDB, and Datadog, which generate revenue based on actual consumption and data throughput, have seen their stock prices hit peak levels. They are viewed as the infrastructure that AI agents will use. Conversely, companies like Asana, Monday.com, and Workday, which rely on the number of human seats, are trading near their lows. As AI agents replace human workers, the number of seats declines, directly cannibalizing the revenue of any company that charges per user.

Survival in the next era of software is no longer about providing a convenient interface for humans to enter data. It is about owning the data that allows an AI agent to perform the work. The transition from a tool-based economy to an outcome-based economy means that the software is no longer the product; the proprietary data that powers the agent is the product.

The collapse of SaaS multiples proves that the market no longer believes in the permanence of the software interface. When the value shifts from the tool to the intelligence, only those with exclusive data and flexible pricing will remain.