Not too long ago, a founder sent a text to his investor with news: he was swapping out his whole customer service staff for Claude Code, an AI solution capable of independently writing and deploying software. To Lex Zhao, an investor at One Way Ventures, this message signified a larger shift — the point at which enterprises like Salesforce ceased to be the automatic choice.
“The hurdles to software creation are now so minimal due to coding agents that the build versus buy decision is increasingly favoring building in numerous instances,” Zhao stated to TechCrunch.
The shift from build to buy is merely one aspect of the issue. The entire concept of employing AI agents instead of human workers raises questions about the SaaS business model itself. SaaS firms typically charge for their software per seat, based on how many employees log in to utilize it. “SaaS has long been considered one of the most appealing business models because of its highly reliable recurring revenue, extensive scalability, and gross margins of 70-90%,” Abdul Abdirahman, an investor at the venture capital firm F-Prime, remarked to TechCrunch.
When one or a few AI agents can perform that labor — when employees simply instruct their preferred AI to extract the data from the system — the per-seat pricing model begins to falter.
The swift development of AI also implies that new tools, such as Claude Code or OpenAI’s Codex, can mirror not only the essential functionalities of SaaS products but also the supplementary tools a SaaS provider might sell to enhance revenue from current clients.
Additionally, customers now possess the ultimate negotiation tool: If they object to the pricing of a SaaS provider, they can more easily than ever craft their own alternative. “Even if they opt not to pursue the build route, this exerts downward pressure on the contracts that SaaS vendors can secure during renewals,” Abdirahman added.
This trend was evident as early as late 2024, when Klarna revealed it had abandoned Salesforce’s primary CRM product for its own internally developed AI system. The dawning realization that an increasing number of other firms can emulate this is unsettling public markets, where the stock values of SaaS giants such as Salesforce and Workday have been declining. In early February, an investor sell-off erased almost $1 trillion in market value from software and services stocks, followed by another billion loss later that month.
Techcrunch event
San Francisco, CA
|
October 13-15, 2026
Analysts are referring to this phenomenon as the SaaSpocalypse, with one noting it as FOBO investing — or the fear of obsolescence.
Nonetheless, the venture investors that TechCrunch consulted believe such anxieties are merely temporary. “This isn’t the end of SaaS,” Aaron Holiday, a managing partner at 645 Ventures, stated. Instead, it marks the start of an established entity shedding its skin, he suggested.
Act swiftly, disrupt SaaS
The trend in the public market is best exemplified by Anthropic’s recent product rollouts. The company launched Claude Code for cybersecurity, and relevant stocks fell. It introduced legal tools in Claude Cowork AI, and the stock price of the iShares Expanded Tech-Software Sector ETF — a collection of publicly traded software companies including LegalZoom and RELX — also declined.
In some respects, this was anticipated, as SaaS firms had long been overpriced, according to investors. It also doesn’t alleviate the situation that these organizations experienced most of their growth during the era of zero-interest rates, which has now concluded. Operating costs increase when borrowing expenses rise.
Typically, public market investors evaluate SaaS companies by forecasting future revenue. However, it is uncertain whether in a year or five, anyone will use SaaS products to the extent they once did. That’s why every time a new sophisticated AI tool is launched, SaaS stocks react with tremors.
“This could be the first instance in history where the terminal value of software is fundamentally challenged, significantly altering how SaaS companies are valued moving forward,” Abdirahman noted.
This is because merely adding AI features to existing SaaS products might not suffice. A surge of AI-native startups is emerging rapidly, having entirely redefined what it means to be a software company.
Software development has now become easier and more affordable, making it simpler to replicate, Yoni Rechtman, a partner at Slow Ventures, mentioned to TechCrunch.
This is advantageous for the next wave of startups but detrimental for the established players that invested years in developing their technology stacks.
Conversely, the market currently lacks enough time and evidence to demonstrate that whatever new business model arises in the aftermath of SaaS will be beneficial. AI firms are sometimes structuring their pricing models based on consumption, meaning customers pay according to their AI usage, quantified in tokens (with each provider defining this slightly differently).
Others are pursuing “outcome-based pricing,” where charges are applied based on how well the AI functions. Ironically, this is the present strategy of former Salesforce CEO Bret Taylor’s AI startup, Sierra, which offers customer service agents and serves as a quasi-competitor to Salesforce.
Thus far, this method appears to be effective. In November, Sierra achieved $100 million in annual recurring revenue within less than two years.
There was once a notion that cloud-based software, such as SaaS, wouldn’t depreciate and could endure for decades. In some respects, this remains true compared to earlier offerings — on-premises software, which organizations had to install and manage on their servers.
However, being cloud-based does not shield SaaS providers from a completely new technology emerging as competition: AI.
Investors are understandably apprehensive as AI-native firms are established, adapt, and develop technology at a pace that traditional SaaS companies cannot match. SaaS corporations are, after all, the incumbents, having replaced the bygone on-premises vendors during the previous era of disruption.
This SaaSpocalypse evokes a Taylor Swift lyric about what unfolds when “someone else lights up the room” because “people adore a newcomer.”
“The key takeaway from the SaaS pullback is that it’s both a genuine structural shift and possibly a market overreaction,” Abdirahman remarked, adding that investors generally “sell first and pose inquiries later.”
SaaS IPOs are suspended
Public-market SaaS companies aren’t the only entities experiencing investor trepidation.
A Crunchbase report released on Wednesday indicated that, while the IPO landscape seems to be improving for certain sectors, there haven’t been — and are not anticipated to be — any venture-backed SaaS filings on the horizon.
Holiday mentioned that this might stem from substantial pressure on larger, private, late-stage SaaS companies like Canva and Rippling due to the finicky IPO environment, lofty expectations fueled by AI innovations, and the unstable stock prices of already listed SaaS companies.
Some of these companies, among them mid-sized SaaS firms, have even faced challenges in securing extension rounds in the private market, Holiday indicated, due to the same concerns that public investors hold.
“No one wishes to endure the volatility of public markets when sentiment can trigger companies into downward spirals,” Rechtman noted, predicting that such companies will remain private for an extended period.
In the meantime, the public market anticipates gaining insight into the financials of the first AI-native companies aspiring to IPO. Speculation suggests that both OpenAI and Anthropic are considering IPOs, potentially within this year.
The most likely scenario is one that integrates the old and the new, as tech disruptions have historically done.
Holiday stated that many of the new features businesses are exploring “won’t endure” and that enterprises will consistently require software that complies with regulations, supports audits, manages workflows, and ensures durability.
“Sustainable shareholder value isn’t established on hype,” he further explained. “It’s built on fundamentals, retention, profit margins, real budgets, and defensibility.”