The landscape of enterprise technology is undergoing a seismic shift as we navigate through 2026. Software as a Service, once the untouchable titan of efficiency, is facing a “SaaSpocalypse” where the traditional value of code is evaporating in the face of generative AI. To understand this transformation, we are joined by Vijay Raina, a seasoned specialist in enterprise SaaS technology and software architecture. Vijay has spent years advising organizations on how to build robust, scalable digital tools, but today he is witnessing a world where multi-million dollar software contracts are being dismantled by internal teams with nothing more than a well-structured prompt and a few days of effort.
The core themes of this discussion revolve around the total commoditization of software code and the resulting destruction of market value for legacy vendors. We explore how major enterprises are bypassing traditional procurement and tenders to “vibe code” their own solutions, the criteria that now determine whether a software vendor can survive—specifically urgency, immediate value, and “stickiness”—and why the year 2027 will represent a massive reckoning for contract renewals. We also dive into the cultural shift required for Western companies to adapt to this leaner, faster development cycle and why the ability to model business processes has become far more valuable than the ability to write a line of Rust or Python.
Large organizations are now successfully rebuilding expensive, specialized software tools in just a few days using AI. How is this reality fundamentally changing the power dynamic between software vendors and their corporate clients?
The power dynamic has completely flipped because the barrier to entry for creating complex functionality has vanished. For years, vendors held clients hostage with the complexity of their codebases, but we recently saw a French e-commerce giant, La Redoute, take a tool they were paying €360,000 a year for and rebuild it in just 48 hours using AI. They didn’t just copy it; they improved it by tuning the copy to their specific environment, something the original vendor’s rigid architecture would never have allowed. When a company realizes they can replace a third-of-a-million-euro line item with two days of internal work, the software vendor loses all leverage. We are moving into an era where the price of software is no longer fixed by the vendor’s perceived value but is instead a negotiation based on how much it would cost the client to just build a better version themselves over a weekend.
You’ve mentioned the concept of “vibe coding” as a way for non-developers to create complex platforms. Could you describe how this process works in a professional setting and what it means for traditional IT departments?
Vibe coding is essentially the democratization of software architecture through natural language, and it is fascinating to watch it play out in real-time. I’ve seen experts who can’t write a single line of code use LLMs to rework entire front-ends, moving from a stagnant, outdated community platform to a modern, responsive interface in just a few hours of conversation with an AI agent. It’s an iterative, almost sensory process where you talk to the agent, see the results, and refine the “vibe” until it matches your dream requirements. For traditional IT departments, this is a wake-up call; the months-long cycle of requirements definition, technical provisos, and bickering over compromises is being replaced by instant execution. When you can have an agent that handles other agents, running tests and writing detailed prompts on WhatsApp, the old-school developer who focuses only on syntax becomes a bottleneck rather than an asset.
In this new environment, you suggest a three-part test for SaaS vendor survival: urgency, value, and stickiness. Why is stickiness suddenly the most critical factor for a company’s longevity?
Stickiness is the only thing that prevents a customer from looking at your product on Monday and deciding to clone it by Wednesday. In the past, “value” was enough, but today, if your product is highly successful and widespread, it actually becomes a target for replication because the logic is so well-understood. Look at the company Smartway, which used software to track food waste and manage discounts in supermarkets; eventually, big retailers like Carrefour and Intermarché built their own versions because they saw the clear value. Stickiness comes from things that AI cannot easily replicate: a network of approved physical repairers, complex hardware certifications, or access to proprietary external data that the customer needs to operate. If your only moat is the code you’ve written, you are standing on quicksand because that code can now be modeled and reproduced in minutes.
We are seeing a trend where even the largest vendors, like Salesforce and SAP, are having their budgets slashed by retailers. What is the logic behind these companies wanting to buy only the “engine” while building everything else themselves?
The logic is a brutal pursuit of efficiency and a rejection of “bloatware” that has plagued enterprise contracts for decades. Retailers are realizing they have been paying for mandatory modules they never touched and features that never quite fit their specific workflows. Now, they are going to these giants and telling them they only want the core engine—the robust database and processing power—because they can use vibe coding to build the user interfaces and specific functional modules themselves at a fraction of the cost. I’ve seen cases recently where a major brand cut its Salesforce budget to a third, and another retailer halved their contract during renegotiation. They are reclaiming the “last mile” of the software experience, ensuring it perfectly matches their internal processes rather than forcing their employees to adapt to a vendor’s rigid template.
If the value of software is no longer in the code itself, where exactly has that value migrated to in the modern B2B ecosystem?
The value has migrated upstream to the human ability to visualize, model, and project exactly what a business needs. It is no longer about the “how”—the syntax and the deployment—but the “what” and the “why.” Prompting is a skill, but the real expertise lies in process modeling; knowing how a marketing funnel should actually flow or how a logistics chain should respond to a disruption. There is also immense value shifting toward the infrastructure, the networks, and the people who can integrate and secure these AI-generated tools. We are seeing a “buying make” middle ground where companies hire freelancers to build custom clones in two or three days rather than signing a five-year SaaS contract. The value is now in agility and the capacity to express complex business requirements in a way that an AI can execute perfectly.
The “ArtCraft” project recently cloned several Adobe applications using AI and Rust. What does this specific case tell us about the future of professional creative suites and the limits of AI-driven replication?
The ArtCraft project is a fascinating litmus test because it shows how quickly the “look and feel” of professional tools can be commoditized; gaining 40,000 stars on GitHub in a few days shows a massive hunger for alternatives to high subscription fees. However, it also highlights the current limits: while it can open most native .psd files and covers about 76% of basic features, it only handles about 45% of what a high-end professional truly demands. Mature software like Photoshop is built on decades of “edge cases”—those tiny, rare problems that only happen once in a million uses—which are incredibly hard for an AI to model without decades of historical data. So, while the “average” user might move to a free, AI-generated clone, the professional defense for vendors lies in that last 20% of extreme reliability, deep file compatibility, and the legal safety net of a major corporation.
There seems to be a cultural divide between how Chinese and Western retailers view IT and software development. How does this influence their ability to survive the SaaSpocalypse?
It is a profound cultural difference that often shocks Western executives. In China, big retailers don’t see themselves as IT companies; they don’t have the oversized, internal development teams that we see in Europe or the US. They use shared architectures and platforms like Alibaba or Tencent, where everyone gets the same new features simultaneously, and they focus their energy on the physical promise—pricing products right and ensuring 30-minute deliveries. In the West, we’ve spent years building massive, sensitive internal systems that we have to keep busy, which leads to over-engineering. As we enter this AI era, Western companies are finding themselves with huge teams of developers they no longer need to write code, while the Chinese model of focusing on the business outcome rather than the “pipes” is proving to be much more naturally aligned with how AI functions.
Critics argue that while AI can cut the cost of building software, it doesn’t necessarily cut the cost of maintaining it. How do you respond to the concern that these “home-grown” AI tools will eventually become a maintenance nightmare?
The “maintenance nightmare” argument is often a defensive shield used by legacy vendors, but the reality is changing. First, the “build” phase is historically much more expensive and labor-intensive than maintenance; if you eliminate that massive upfront cost, you’ve already won. Second, maintenance in an AI-driven world is more about swapping APIs and prompting for updates rather than manually hunting for bugs in a million lines of legacy code. If a security standard changes, you don’t spend three weeks in a support ticket; you ask the agent to update the environment and run a fresh battery of tests. We are also seeing that token costs—the price of running these AIs—are staying competitive due to the fierce rivalry between players like OpenAI, DeepSeek, and Mistral. The cost of maintenance is becoming an internal cost of agility rather than an external, fixed fee that grows every year.
Gartner is still forecasting an increase in software spending for the next year. Why do you believe their data might be missing the “violent change” you are describing?
Gartner is looking through the rear-view mirror at multi-year contracts that were signed before this AI revolution reached its current velocity. They see the 15% growth because those checks are already written and the budgets are locked in. I am looking at the 2027 renewals—that is where the cliff is. By 2027, companies will have spent the last year running tests, building small internal tools, and gaining the confidence to tell their vendors “no.” They won’t necessarily cancel everything, but they will negotiate with their backs to the wall, demanding that SaaS prices be divided by four because they know the alternative is now feasible. We’ve never seen a world where a “cost center” like IT can suddenly generate positive budget lines by cancelling massive contracts and replacing them with internal agility, and that is a shift no traditional forecast can capture yet.
If a SaaS founder is worried about their product being “cloned” or commoditized, what is the single most important action they should take today to protect their business?
The most dangerous thing a founder can do is rely on their “technological lead” as a shield. Instead, they need to have a brutally honest conversation with their customers right now and ask them exactly why they are paying. You have to move away from being a “code provider” and toward being an “outcome partner.” If your customers are sticking with you, is it because of the interface, or is it because you provide external market data they can’t get elsewhere? Is it because you handle the legal certifications they don’t want to touch? If you can’t answer that, they will copy you. Founders who are already reworking their offers, testing new collaborative models, and acknowledging that their code is no longer the moat are the ones who will survive. You have to look up from the keyboard and start building a new landscape where you bring value through expertise, not just through the software you’ve shipped.
What is your forecast for the B2B software market over the next eighteen months?
I expect we will see a massive “value destruction” in the traditional sense, where billions in market cap will vanish as software becomes negotiable. However, this isn’t the death of the industry; it’s a rebirth. We will see the rise of the “super-freelancer” and small, agile teams that can deploy custom, secure, and integrated enterprise tools in days. The “make or buy” debate will be replaced by a culture of continuous creation, where software is treated as a disposable, evolving tool rather than a static asset. Those who thrive will be the organizations that stop trying to protect their legacy code and instead focus on mastering the art of the prompt and the depth of the business model. The era of the “software vendor” is shifting into the era of the “solutions architect,” and it’s going to be a brutal but necessary reality check for everyone involved.
