Not long ago, a founder texted his investor with an update: he was replacing his entire customer service team with Claude Code, an artificial intelligence tool that can write and develop software by itself. To Lex Zhao, an investor at One Way Ventures, the message pointed to something bigger — a moment when companies like Salesforce stopped being the automatic default.
“The barriers to entry for building software are so low now thanks to coding agents, that the build-versus-buy decision is shifting toward building in so many cases,” Zhao told TechCrunch.
The make vs. buy switch is only part of the problem. The whole idea of using AI agents instead of humans to perform tasks calls into question the very SaaS business model. SaaS companies currently price their software per seat — that is, by how many employees log in to use it. “SaaS has long been seen as one of the most attractive business models due to highly predictable recurring revenue, massive scalability and 70-90% gross margins,” Abdul Abdirahman, an investor at venture capital firm F-Prime, told TechCrunch.
When one or a few AI agents can do this job—when employees simply ask the AI of their choice to pull data from the system—that per-position model begins to break down.
The rapid pace of AI development also means that new tools, such as Claude Code or OpenAI’s Codex, can replicate not only the core functions of SaaS products but also the additional tools that a SaaS vendor would sell to increase revenue from existing customers.
Plus, customers now have the ultimate contract negotiation tool in their pockets: If they don’t like a SaaS vendor’s pricing, they can, more easily than ever, create their own alternative. “Even if they don’t go the build route, that puts downward pressure on the contracts that SaaS vendors can secure during renewals,” Abdirahman continued.
We already saw this at the end of 2024, when Klarna announced that it had abandoned Salesforce’s flagship CRM product in favor of its own AI system. The realization that a growing number of other companies may do the same is spooking public markets, where the share prices of SaaS giants like Salesforce and Workday have tumbled. In early February, an investor sell-off all but disappeared $1 trillion in market value by software and services stocks, followed by another billion later in the month.
Techcrunch event
San Francisco, California
|
13-15 October 2026
Experts call it the SaaSpocalypse, with one analyst dubbing it FOBO investing — or fear of being overcome.
However, venture capitalists that TechCrunch spoke with believe that these fears are only temporary. “This is not the death of SaaS,” Aaron Holiday, managing partner at 645 Ventures, told TechCrunch. Rather, it is the beginning of an old snake shedding its skin, he said.
Move fast, break SaaS
THE public market pattern best demonstrated by Anthropic’s recent product launches. The company released the Claude code for cyber security, and its shares fell. It launched legal tools at Claude Cowork AI, and the share price of the iShares Expanded Tech-Software Sector ETF — a basket of publicly traded software companies that includes the likes of LegalZoom and RELX — also fell.
In some ways, that was to be expected, as SaaS companies have long been overvalued, investors said. It also doesn’t help that these companies did most of their growth during the era of zero interest rates, which has since ended. The cost of doing business increases when the cost of borrowing money increases.
Public market investors typically price SaaS companies by estimating future revenue. But there’s no telling whether in a year or five years anyone will be using SaaS products to the extent they once did. That’s why every time a new advanced AI tool is released, SaaS stocks feel the jitters.
“This may be the first time in history that the terminal value of software is fundamentally challenged, fundamentally reshaping how SaaS companies are undertaken in the future,” said Abdirahman.
This is because layering AI features on top of existing SaaS products may not be enough. A horde of AI startups is growing at a record pace, having completely redefined what it means to be a software company.
Software is now easier and cheaper to build, which means it’s easier to replicate, Yoni Rechtman, a partner at Slow Ventures, told TechCrunch.
That’s good news for the next generation of startups, but bad news for incumbents who spent years building their technology stacks.
On the other hand, the market also lacks enough time and evidence to show that any new business model that emerges after SaaS will be worthwhile. AI companies sometimes price their models on a consumption basis, meaning customers pay based on how much AI they use, measured in tokens (which each model provider defines slightly differently).
Others are working on “performance-based pricing,” where fees are charged based on how well the AI actually works. This, ironically, is the current approach of former Salesforce CEO Bret Taylor’s AI startup, Sierra. quasi competitor to Salesforce which offers customer service agents.
The approach seems, so far, to be working. In November, Sierra hit $100 million in annual recurring revenue in less than two years.
There was also once the idea that cloud-based software like SaaS sells will never depreciate and could last for decades. This is still somewhat the case compared to what came before it — on-premise software, which companies had to install and maintain on their own servers.
But being in the cloud doesn’t protect SaaS vendors from a whole new technology rising to compete: artificial intelligence.
Investors are rightfully nervous as native AI companies emerge, adapt, adopt and build technology much faster than a traditional SaaS company can move. SaaS companies are, after all, the incumbents themselves, having replaced old-school on-premises vendors in the latest era of disruption.
This SaaSpocalypse brings to mind that Taylor Swift lyric about what happens when “someone else lights up the room” because “people love an ingénue.”
“The most important thing to understand about SaaS retirement is that it’s both a real structural change and probably a market overreaction,” Abdirahman said, adding that investors typically “sell first and ask questions later.”
SaaS IPOs are on hold
Publicly traded SaaS companies aren’t the only ones getting chills from investors.
However, a Crunchbase report released Wednesday showed that the IPO market appears to be thawing for some sectorsthere haven’t been – and aren’t expected to be – any venture-backed SaaS filings on the horizon.
Holiday said this may be because there is a lot of pressure on large, private SaaS companies like Canva and Rippling, given the persnickety IPO, high expectations driven by AI developments and the volatile stock price of already public SaaS companies.
Some of those companies, including mid-sized SaaS companies, have even struggled to raise expansion rounds in the private market, Holiday said, with the same fears that public investors have.
“No one wants to be subject to the volatility of the public markets when the climate can send companies into downward spirals,” Rechtman said, adding that he expects to see companies like these remain private for much longer.
Meanwhile, the public market is waiting to get a good look at the financials of the first native AI companies hoping for an IPO. The scuttlebutt says that both OpenAI and Anthropic are considering IPOs, perhaps even later this year.
The most likely outcome is something that combines the old and the new, as technological disruptions always do.
Holiday said most of the new features companies are playing these days “won’t stick” and that businesses will always need software that meets compliance regulations, supports audits, manages workflow and offers resiliency.
“Sustained shareholder value is not based on hype,” he continued. “It’s built on fundamentals, retention, margins, real budgets and defensibility.”
