A recent incident saw a founder inform his investor that his entire customer service operation was being replaced by Claude Code, an AI tool capable of independently writing and deploying software. For Lex Zhao, an investor at One Way Ventures, this communication signified a profound shift: the end of an era where companies like Salesforce were the unquestioned standard.
“The entry barriers for software development have become remarkably low due to the advent of coding agents, causing the traditional 'build versus buy' dilemma to increasingly favor the 'build' option,” Zhao explained to TechCrunch.
This reevaluation of "build versus buy" is merely one facet of a broader challenge. The fundamental concept of deploying AI agents to execute tasks traditionally performed by humans directly undermines the core SaaS business model. SaaS providers typically license their software on a per-seat basis, charging according to the number of employee users. Abdul Abdirahman, an investor at F-Prime, highlighted to TechCrunch that “SaaS has long been considered one of the most appealing business models due to its highly predictable recurring revenue, substantial scalability, and impressive 70-90% gross margins.”
However, when a single AI agent, or a small group of them, can accomplish the same workload – with employees merely instructing their preferred AI to retrieve system data – the viability of this per-seat pricing model begins to erode.
Furthermore, the accelerating pace of AI innovation implies that emerging tools, such as Claude Code or OpenAI’s Codex, are capable of replicating not only the primary functionalities of SaaS offerings but also the supplemental tools that vendors traditionally sell to expand revenue from their current client base.
Adding to this, customers now possess a formidable negotiation tool: if dissatisfied with a SaaS vendor's pricing, they can, with unprecedented ease, develop their own bespoke solutions. Abdirahman elaborated, “Even if they do not take the build route, this creates downward pressure on contracts that SaaS vendors can secure during renewals.”
This trend was evidenced as early as late 2024, when Klarna publicly announced its decision to abandon Salesforce’s primary CRM product in favor of an internally developed AI system. The growing awareness that an increasing number of enterprises can emulate this move is unsettling public markets, leading to declining stock prices for SaaS behemoths such as Salesforce and Workday. Notably, an investor sell-off in early February eradicated nearly $1 trillion in market value from software and services stocks, with an additional billion lost later that month.
Industry experts have coined this phenomenon the “SaaSpocalypse,” with one analyst labeling it “FOBO investing”—signifying the fear of becoming obsolete.
However, venture investors interviewed by TechCrunch largely consider these anxieties to be transient. Aaron Holiday, a managing partner at 645 Ventures, assured TechCrunch, “This isn’t the death of SaaS.” Instead, he characterized it as a period of profound transformation, akin to an old snake shedding its skin.
The prevailing public market trend is vividly demonstrated by Anthropic’s recent product introductions. Following the release of Claude Code for cybersecurity, corresponding stocks experienced declines. Similarly, when the company launched legal tools within Claude Cowork AI, the stock price of the iShares Expanded Tech-Software Sector ETF – an index comprising publicly traded software firms such as LegalZoom and RELX – also saw a decrease.
In certain respects, this market reaction was anticipated, as investors indicated that SaaS companies had, for an extended period, been overvalued. Compounding this, a significant portion of these companies' growth occurred during the zero-interest-rate environment, which has now concluded. Consequently, the rising cost of borrowing capital inherently increases the overall cost of doing business.
Public market investors traditionally value SaaS companies by forecasting future revenue. However, the extent to which SaaS products will be utilized in one or five years is now uncertain. This fundamental ambiguity explains why each launch of a new, advanced AI tool invariably sends ripples through SaaS stock valuations.
“This may be the first time in history that the terminal value of software is being fundamentally questioned, materially reshaping how SaaS companies are underwritten going forward,” Abdirahman asserted.
This uncertainty stems from the fact that merely integrating AI features into existing SaaS products may prove insufficient. A proliferation of AI-native startups is emerging at an unprecedented rate, fundamentally redefining the very essence of a software company.
Yoni Rechtman, a partner at Slow Ventures, informed TechCrunch that software is now both simpler and more economical to develop, which inherently makes it easier to replicate.
While this represents a favorable development for the next wave of startups, it poses a significant challenge for incumbent companies that have invested years in constructing their proprietary tech stacks.
Conversely, the market currently lacks sufficient time and empirical data to conclusively demonstrate the efficacy of any new business models arising in the wake of SaaS. Some AI companies are adopting consumption-based pricing, where clients are billed according to their AI usage, typically quantified in tokens (a metric that varies slightly among model providers).
Other innovators are exploring “outcome-based pricing,” a model where fees are contingent upon the actual performance and effectiveness of the AI. Ironically, this is the strategy currently employed by Sierra, the AI startup founded by former Salesforce CEO Bret Taylor, which functions as a quasi-competitor to Salesforce by providing customer service agents.
This innovative approach appears to be yielding positive results thus far. By November, Sierra had achieved an impressive $100 million in annual recurring revenue in less than two years since its inception.
Previously, there was a widespread belief that cloud-based software, such as SaaS offerings, would exhibit perpetual value and remain relevant for decades. This notion largely holds true when contrasted with its predecessor, on-premises software, which necessitated installation and maintenance on proprietary servers.
Nevertheless, merely residing in the cloud does not insulate SaaS vendors from the competitive challenge posed by an entirely new and ascendant technology: Artificial Intelligence.
Investors are justifiably apprehensive as AI-native companies rapidly emerge, adapt, integrate, and develop technology at a pace significantly exceeding that of traditional SaaS enterprises. It is crucial to remember that SaaS companies are themselves the current incumbents, having superseded old-school on-premises vendors in the preceding wave of technological disruption.
This "SaaSpocalypse" evokes a particular Taylor Swift lyric, highlighting the dynamic when “someone else lights up the room” and “people love an ingénue.”
“The most important thing to understand about the SaaS pullback is that it is simultaneously a real structural shift and potentially a market overreaction,” Abdirahman stated, further noting that investors typically “sell first and ask questions later.”
It is not solely public-market SaaS companies that are sensing a cooling in investor sentiment.
A recent Crunchbase report revealed that, while the IPO market appears to be thawing for certain sectors, there have been no venture-backed SaaS filings on the horizon, nor are any anticipated.
Holiday attributed this situation to significant pressure on large, private, late-stage SaaS companies, such as Canva and Rippling. This pressure stems from a challenging IPO window, elevated expectations fueled by AI advancements, and the volatile stock performance of already public SaaS entities.
Holiday further noted that some of these companies, including mid-size SaaS firms, have encountered difficulties in securing extension rounds within the private market, driven by the identical concerns held by public investors.
“Nobody wants to be subjected to the volatility of public markets when sentiment can send companies into downward tailspins,” Rechtman commented, expressing his expectation that such companies will opt to remain private for an extended duration.
Concurrently, the public market keenly awaits an examination of the financials from the initial wave of AI-native companies contemplating IPOs. Speculation suggests that both OpenAI and Anthropic are considering public listings, possibly as early as later this year.
The most probable outcome, consistent with historical technological disruptions, is an integration that seamlessly blends existing and nascent innovations.
Holiday posited that many of the novel features companies are currently experimenting with “won’t stick.” He emphasized that enterprises will consistently require software that adheres to compliance regulations, facilitates audits, streamlines workflow management, and provides robust durability.
He concluded, “Durable shareholder value isn’t built on hype. It’s built on fundamentals, retention, margins, real budgets, and defensibility.”
The Editorial Staff at AIChief is a team of professional content writers with extensive experience in AI and marketing. Founded in 2025, AIChief has quickly grown into the largest free AI resource hub in the industry.