The future of SaaS is in question. At the start of 2026, a mass market selloff erased $1 trillion in market capitalization from software stocks and raised concerns that AI agents could replace software workflows. Now, just months later, many are arguing that the worst of the SaaSpocalypse is behind us. In June, Business Insider executive editor Joe Ciolli released an article arguing that the SaaSpocalypse is “dead,” citing a 13% increase in software stocks across the iShares Expanded Tech-Software ETF. Similarly, Thoma Bravo founder and managing partner Orlando Bravo reportedly told CNBC that the SaaSpocalypse is over and argued that “AI is an enormous tailwind for software companies.”

While the SaaS market is showing signs of improvement, AI coding tools like Claude Code and Codex present a significant challenge to software vendors. On the one hand, narrow vendors with thin user interfaces run the risk of being displaced by coding tools and on the other, complex platforms remain well placed to survive, especially if they leverage AI to build agentic experiences.

What is clear is that the way software is made is changing. Software vendors now operate in a reality where anyone can spin up an app in a matter of minutes. Whether that software is fit for deployment at scale is another matter, but in any case, the barrier to entry for software production is lower than ever before. This means SaaS vendors need to look for new ways to provide value.

How Real Is The SaaSpocalypse?

At the start of the year, the SaaSpocalypse made headlines due to rapid stock sell offs, but as the dust settles, many in the industry argue that AI will uplift SaaS as much as disrupt it.

In April, Salesforce CEO Marc Benioff highlighted the upside in AI tools, telling the Wall Street Journal, “People think we have our back against the wall when in fact the opportunity has never been greater.”

So just how tangible is the SaaSpocalypse today? “In the stock market, it’s absolutely a real thing…stock prices are going down, however, I don’t think that it is real in the way that it’s actually going to impact us long term,” Terra Higginson, principal research director at Info-Tech Research Group, told me at Info-Tech LIVE 2026 in Las Vegas in June.

Higginson said that there’s a lot of fear that tools are going to become unnecessary and that companies are going to be able to build everything, but said, “That’s not true.” However, she does note concerns for thinner vendors, specifically SaaS products that just do one thing.

“Those guys are in trouble right now,” Higginson said. “They don’t own a lot of the workflow, they don’t own a lot of the transaction, they don’t own a lot of the data and they’re easy to replicate through vibe coding.” Such vendors risk being swept up by the SaaSpocalypse.

From a top down perspective, larger platforms like Salesforce, Microsoft and SAP are likely to be more difficult to replace, not just due to their widespread adoption, but also the underlying product ecosystem, data sources, compliance and governance capabilities. In addition, all of these vendors can leverage AI agents to accelerate product development internally.

“The ability to generate code quickly does not replace the role enterprise software plays within large organizations. Companies are not simply purchasing features; they are investing in reliable systems that are embedded in workflows and designed to operate at scale," Ray Shu, senior managing director and head of originations, technology media and telecom banking at Capital One, told me via email.

Salesforce And The Evolution of SaaS

Alongside concerns surrounding the SaaSpocalypse, enterprise software creation itself is also undergoing a monumental shift, not just disrupted by AI coding, but also by the need to provide access to autonomous agents. After all, today’s apps aren’t just being built with the help of coding agents; they’re also being designed for agents.

Madhav Thattai, executive vice president and general manager of Agentforce at Salesforce, told me in a video interview that over the past three years, the company has not only thought about ways to bring generative capabilities into applications, but also how to create agentic experiences for customers. He also says SaaS apps are going to evolve so that agents become “deeply integrated” into those experiences.

“We believe that we will live in a world where humans and agents are orchestrating, collaborating and driving incredible experiences together. And so we’ve designed the software from the beginning to optimize for both,” Thattai said.

The company has modernized its core CRM platform into what its website describes as an “agentic CRM”, bringing together people, agents, apps and data under one platform. The idea is to provide enterprises with the tools necessary to build agentic experiences. Agentforce is one such tool, providing a platform to create and customize autonomous agents.

Thattai also highlights there’s a difference between building an app and providing value. “You hear a lot about vibe coding and all of that, and people can build anything quickly and ship it, and you could build a good demo, but that is not where outcomes are delivered,” Thattai said. What matters most, he argues, is driving outcomes for customers, giving them the tools and capabilities to optimize those experiences.

From this perspective, large SaaS vendors like Salesforce remain relevant in the age of vibe coding not just because of widespread adoption among enterprise customers, but also due to observability, analytics, optimization and compliance capabilities that are difficult to replicate with a vibe-coded app.

Consolidating SaaS Platforms

Another challenge facing the SaaS industry is that of sprawl. According to research from AI-powered SaaS management provider BetterCloud, organizations use an average of 106 different SaaS tools. This high number of apps is not only difficult to manage but is also creating a need for consolidation in the industry.

“Software sprawl rarely happens by accident. It happens one reasonable decision at a time; a team adopts a new tool to solve a real problem, and a year later you have ten tools solving overlapping problems and nobody owns the full picture,” Krishna Sai, chief technology officer at cybersecurity vendor SolarWinds, told me via email.

"The future of SaaS will be determined by the value vendors deliver, not the hype they generate. Times of market consolidation, like the one we’re seeing right now, tend to separate organizations with sustainable products from those built primarily on marketing momentum,” Sai said.

During this period of consolidation, SaaS vendors will face increased pressure to provide value. “Customers are scrutinizing every renewal now, and they’re asking a very simple question: what did this actually do for us? Hype doesn't survive that question. Value does."

For Sai, consolidation isn’t simply about reducing the number of tools, but asking whether the technology stack can keep up with how fast the business is changing. “My guidance every time is to stop asking what a tool does and start asking what it’s worth,” he said.

As software becomes easier to create with AI coding, the dynamics of the buy vs. build debate are changing too. AI has lowered the barrier to entry for creating software, giving companies the opportunity to experiment and build working prototypes faster than with manual coding.

“AI has changed the economics of software creation. Tasks that once required large teams can now be prototyped much faster, which naturally reopens the buy-versus-build debate,” Andrea Malagodi, CTO at Sonar, told me via email.

“SaaS vendors now have to prove their value beyond the interface. If natural language becomes the front end for more workflows, the real moat is no longer just usability. It’s the depth of the product, the quality of the outcomes, the trustworthiness of the system and how well it fits into an enterprise environment,” Malagodi said.

However, while building software is becoming easier, maintenance remains a challenge. “Creating an application is trivial compared to maintaining it through time," Jeremiah Stone, CTO of SnapLogic, told me via email. “Vibe coding a first version of an app is a down payment on a long-term commitment and this long-term commitment is the true reason why software vendors exist.”

Stone added that it’s not a great use of capital to allocate staff to build apps that are non-differentiating. Similarly, Malagodi notes the question isn’t “can we build this”; it’s “can we build, secure, maintain and evolve this responsibly over time?”

In this sense, companies not only need to consider the upfront time investment of building an app from scratch, but also the long-term maintenance. Vibe coding offers a fast way to build a working application, but such apps are unlikely to be able to compete with the scale of features and governance offered by entrenched SaaS vendors.