One day not too long ago, our founder texted an update to his investors. We’re replacing our entire customer service team with Claude Code, an AI tool that lets you write and deploy software on your own. For Lex Zhao, an investor at One Way Ventures, the message signaled something bigger: a moment when companies like Salesforce stopped automatically becoming the default.
“The barrier to entry to creating software is now so low thanks to coding agents that in many cases, the build-versus-buy decision has shifted to building,” Zhao told TechCrunch.
Changes in building and buying are only part of the problem. The whole idea of using AI agents to perform tasks instead of humans calls into question the SaaS business model itself. SaaS companies currently set a per-seat price for their software. This refers to the number of employees who log in to use the software. “SaaS has been considered one of the most attractive business models due to its highly predictable recurring revenue, immense scalability, and 70-90% gross margins,” Abdul Abdirahman, an investor at venture firm F-Prime, told TechCrunch.
The seat-by-seat model begins to break down when one or a small number of AI agents can do the work—when employees simply ask the AI of their choice to retrieve data from the system.
The fast pace of AI development means new tools like Claude Code and OpenAI’s Codex can replicate not only the core functionality of SaaS products, but also the add-on tools that SaaS vendors sell to increase revenue from existing customers.
Plus, customers have the ultimate contract negotiation tool in their pocket. If you don’t like the pricing of a SaaS vendor, it’s easier than ever to build your own alternative. “Even if you don’t take the build route, this puts downward pressure on the contracts SaaS vendors are able to secure at renewal time,” Abdirahman continued.
We saw this in late 2024 when Klarna announced it was ditching Salesforce’s flagship CRM product in favor of a homegrown AI system. The realization that many other companies could do something similar has spooked public markets, where shares of SaaS giants like Salesforce and Workday have fallen. In early February, investor selling wiped out nearly $1 trillion in software and services stocks’ market capitalization, followed by another $1 billion later that month.
tech crunch event
San Francisco, California
|
October 13-15, 2026
Experts call this the SaaSpocalypse, and some analysts call it FOBO investing, or the fear of becoming obsolete.
But venture investors TechCrunch spoke to believe these concerns are only temporary. “This is not the death of SaaS,” Aaron Holiday, managing partner at 645 Ventures, told TechCrunch. Rather, he said, it’s the beginning of an older snake shedding its skin.
Act fast and beat SaaS
The public market pattern is best illustrated through Anthropic’s recent product launches. The company released “Claude Code” for cybersecurity, and related stocks fell. The company released Claude Cowork AI legal tools, and shares of the iShares Expanded Tech-Software Sector ETF, a basket of publicly traded software companies that includes companies like LegalZoom and RELX, also fell.
Investors say this was partly to be expected, as SaaS companies have long been overvalued. It also doesn’t help that these companies did most of their growth during the zero interest rate era, which has since ended. As borrowing costs rise, so do business costs.
Public market investors typically price SaaS companies by estimating future earnings. But we don’t know if people will be using SaaS products the way they used to in a year or even five years. That’s why SaaS stocks oscillate every time a new advanced AI tool emerges.
“This may be the first time in history that the ultimate value of software is being fundamentally questioned, leading to major changes in how SaaS companies are underwritten going forward,” Abdirahman said.
Because simply adding AI capabilities to an existing SaaS product may not be enough. A swarm of AI-native startups are emerging at a record pace, completely redefining what it means to be a software company.
Yoni Rechtman, partner at Slow Ventures, told TechCrunch.
This is good news for the next generation of startups, but bad news for established companies that have spent years building their technology stacks.
On the other hand, the market lacks enough time and evidence to show that whatever new business model emerges after SaaS is worthwhile. AI companies may price their models based on consumption. That is, customers pay based on their AI usage measured in tokens (defined slightly differently by model provider).
Some companies are working on “outcome-based pricing,” where they charge based on how well their AI actually performs. Ironically, this is the current approach of former Salesforce CEO Brett Taylor’s AI startup Sierra. Sierra is a quasi-competitor of Salesforce that provides customer service agents.
This approach seems to be working so far. In November, Sierra reached $100 million in annual recurring revenue in less than two years.
There was once a belief that cloud-based software, such as that sold by SaaS, would never depreciate and could last for decades. To some extent, this is still true compared to previous on-premises software, which businesses had to install and maintain on their own servers.
But being in the cloud doesn’t protect SaaS vendors from an entirely new technology emerging to compete: AI.
Investors are understandably nervous as AI-native companies emerge, adapt, adopt, and build technology much faster than traditional SaaS companies can migrate. After all, SaaS companies are themselves incumbents that replaced legacy on-premises vendors during the last era of disruption.
This SaaSpocalypse reminds me of Taylor Swift’s lyrics about what happens when “someone else lights up the room” because “people love original things.”
“The most important thing to understand about a SaaS exit is that it is both a real structural change and a potential market overreaction,” Abdirahman said, adding that investors typically “sell first and ask questions later.”
SaaS IPO pending
Public market SaaS companies aren’t the only ones feeling the chill from investors.
Although the IPO market appears to be thawing in some sectors, no venture-backed SaaS filings are imminent or expected to be filed, according to a Crunchbase report released Wednesday.
Holliday said this may be because large private SaaS companies like Canva and Ripling are under a lot of pressure, given the rigors of IPOs, high expectations due to advances in AI, and volatile stock prices for SaaS companies that are already public.
Some of these companies, including mid-sized SaaS companies, are struggling to even raise extended rounds in the private markets because of the same concerns as retail investors, Holliday said.
“No one wants to be exposed to public market volatility when sentiment could send a company into a downturn,” Rechtman said, adding that such companies are expected to remain private for a long time.
Meanwhile, the public market is waiting to take a closer look at the financials of the first AI-native companies aiming for an IPO. According to scuttlebutt, OpenAI and Anthropic are both considering IPOs, possibly as soon as later this year.
The most likely outcome, as technological disruption always does, is one that interweaves the old with the new.
Holliday said most of the new features companies are considering these days are “not here to stay,” and that companies will always need software that meets compliance regulations, supports audits, manages workflow and provides durability.
“Enduring shareholder value is not built on hype,” he continued. “It’s built on fundamentals, retention, profit margins, real budget and defense.”
