The “SaaSpocalypse” is here. The rise of agentic coding has made building software faster and easier than ever, and it’s changing the way we think about software-as-a-service (SaaS). Software stocks plummeted by $1.6 trillion at the beginning of the year, sparking concerns around the viability of SaaS. But as Holden Spaht, managing partner at software-focused private equity firm Thoma Bravo, explained in a recent LinkedIn post, “AI disruption is real and profound, but not all software is equally exposed to the downside risk.”

We’re at a critical inflection point. The question at the core of this shift is: Why should businesses still opt in to expensive software contracts and take on the onerous task of adopting an off-the-shelf product when they can just “vibe code” a customized solution themselves?

Answering this question is deceptively complex. Yes, agentic coding is a breakthrough tool for certain use cases. No, it is not the right fit in every instance. Yes, agentic and vibe coding will change the way SaaS vendors prototype, build, price, and sell their products. No, it won’t eliminate SaaS as a whole, but some vendors and customers are indeed at risk.

Over the next few years, many businesses will attempt to use agentic coding tools to build their own solutions, some will succeed and many will fail. I suspect we will hear more about successes. Software vendors will be forced to take a hard look at their cost to ship, business models, and offerings. AI will continue to permeate our lives on both a professional and personal level. It’s going to be a wild (and potentially volatile) ride. 

Here’s how the SaaSpocalypse is on track to impact businesses, SaaS companies, and consumers. Some of the changes on the horizon may surprise you. 

Businesses: AI is a powerful tool, but think before you build

As we venture deeper into the agentic coding era, some businesses will choose to stick with their existing software vendors, and others will replace them with do-it-yourself (DIY) solutions. While a portion of organizations who chose to build rather than buy will succeed, a significant number will experience remorse around their decision to build software in-house. 

Why? Because companies are in the business of running their business – not building and managing software. Standing up your own software is one thing (and AI makes this easier, no doubt), but maintaining it over time is a completely different animal. As requirements change and bugs emerge, enterprises can quickly find themselves drowning in patchwork fixes, technical debt, and the unrelenting burden of maintaining systems that were never designed to grow and scale with the business.

To be clear, there are domain areas where companies are already building their own software and it’s having a net positive impact – like domain-specific, customer-facing experiences or unique data products. But when it comes to infrastructure and administrative, line-of-business software, vibe coding your own solution isn’t going to be the right decision in 90% of cases. 

The idea of building software to your company’s unique requirements can be alluring. But custom solutions are void of one crucial thing: experience. SaaS lets organizations take advantage of the “wisdom of the crowd” and benefit from an aggregation of other companies’ needs and hard lessons learned. Yes, this can also result in paying for unnecessary features and functionality, but the tradeoff is almost always worth it for the stability and continuous improvement shaped by real-world use. A broad usage base subsidizes learning and accumulation of domain wisdom as much as it does building.

Software vendors: It’s time to rethink your business model and offerings

It’s an extremely exciting time to be in SaaS. Agentic coding is unlocking new opportunities, similar to how the emergence of cloud computing has lowered the barrier to building and scaling software. By enabling a greater variety of individual innovators to create and companies to scale faster, AI is firing up competition and creating more diversity in the marketplace. Not every SaaS company will survive in this rapidly evolving landscape. 

Narrow, niche applications that offer simple processes and tools are going to have a difficult time staying relevant. Hollowed-out SaaS companies that haven’t created a high-context, high-value product are also at risk. This is because the formulaic approach to building an enterprise SaaS company (i.e., finding a workflow, creating a solid UI, and then selling) is no longer going to cut it.

Furthermore, AI is now being used to vet software vendors and determine their true valuation. Private equity firms and VCs are increasingly hiring newly emergent services firms to “reverse build” SaaS offerings via agentic coding. If a company’s product and scope can be easily replicated, why would any organization go through the trouble of acquiring them? We can expect to see more sinking valuations due to this new phenomenon. 

The companies that will thrive in this new environment are those that focus on high-context products encased in a larger service envelope. They understand the importance of providing value that extends beyond the software itself. This can take the form of additional business services, discrete business process outsourcing, and the like. 

Additionally, they recognize the need to alter their pricing structure for the agentic coding era. For example, Intercom’s Fin AI customer agent charges customers on a pay-per-resolution versus a pay-per-seat model. More companies will adopt value-based pricing models as they attempt to better align cost with outcomes and account for the variable, usage-driven costs of AI agents.

Consumers: Bring your own AI (BYOAI)

Companies understand the importance of empowering employees with AI. In response, they’ve rolled out enterprise-grade AI tools (like Gemini and Copilot) with robust, built-in guardrails to boost productivity while keeping their organization’s data safe. Despite this, people often default to what’s most familiar.

Many employees “bring” their own AI tools into the workplace to vibe code customized personal assistants. This shadow AI is creating serious security risks, similar to the early days of bring-your-own-device (BYOD). Instead of putting strict rules around which AI tools employees can and cannot use, companies should instead focus on helping employees use their AI of choice safely and securely. 

Despite its name, the “SaaSpocalypse” isn’t doom and gloom. It represents an inevitable and exciting shift into the next generation of how we work and build software. The vendors that will emerge unscathed understand success is about more than just creating a high-context product – the services surrounding it are equally important. SaaS isn’t dying, the outdated approach to SaaS is.