Stop with the SaaSpocalypse Nonsense Already! Get Real About Vibe-Coding

Observations

The Anatomy of the Hype

The narrative of a SaaSpocalypse reached a fever pitch in early 2026, spurred by market volatility and provocative claims that autonomous AI agents would make the application layer redundant.¹ This serve-two-masters narrative fuels AI-native startups looking to disrupt established moats and Tier 1 consultancies eager to sell multi-million-dollar transformation programmes to save enterprises from obsolescence.

In reality, the February 2026 software sell-off was an indiscriminate repricing of overvalued growth stocks, not a fundamental failure of the SaaS model.² The disruption is not an extinction event; it is a compression of the vendor stack.

SaaS is Far From Dead: The Evidence

The claim that SaaS is dying is mathematically illiterate and provably false in both the short and long term.

In the short term, despite the apocalypse rhetoric, major platforms continue to show robust growth and high retention:

• ServiceNow: Reported 21 per cent year-over-year (Y/Y) subscription revenue growth in Q4 2025, reaching $3.466 billion, with a 98 per cent renewal rate.³
• Salesforce: Achieved $41.5 billion in revenue for FY26 (up 10 per cent Y/Y) and ended the year with a record $72.4 billion in remaining performance obligations (RPO).⁴
• Workday: Projected 2026 financial year subscription revenue of $8.8 billion, representing 14 per cent growth.

The list of usage growth of all the major SaaS platforms continues, and their revenues continue to rise steadily and largely predictably. While the stock market shudders due to external events, SaaS marches on. On top of this, all tier 1 SaaS business solutions (Salesforce, SAP, Dynamics, Oracle, etc.) and most tier 2 SaaS business solutions (TechnologyOne, Zoho, etc.) are embedding AI services deeply into their platforms.

Global SaaS spending is projected to surpass $1 trillion by 2027.⁵ These are not the vital signs of a dying industry; they are the markers of a maturing infrastructure.

The Dangers of Vibe Coding at Enterprise Levels

At the heart of the claims for a SaaSpocalyse is the concept of vibe coding. Using agentic AI to generate entire applications based on conversational prompts. While incredibly valuable for rapid prototyping and for build-to-point solutions, it is dangerous for enterprise core solutions.

• Security and Provenance: Vibe-coded applications often inherit vulnerabilities from training data, including SQL injection and hard-coded secrets, without the oversight of traditional DevSecOps. Prompt injection has even been shown to reveal highly sensitive API and encryption keys from a vibe coder’s desktop. None of the AI coding tool suppliers has a solution to this, other than warning users that vibe coding “has security risks.” That is not good enough if you are building a solution that will run your entire company. For ANZ organisations governed by APRA or CPS 230, vibe coding failures may pass an auditability test, but the very use of the tools may already have broken governance.
• Operational Fragility: Just as with traditional bespoke development, vibe-coded solutions quickly accumulate technical debt. While these solutions provide robust technical documentation, they remain difficult to maintain once the original prompter leaves the organisation. We have learned, over and over again, in the ICT industry, that standardisation of core platforms provides stability and continuity. Vibe-coding scales up this problem.
• Uniformity: Applications developed by vibe-code piecemeal lack uniform design principles, adherence to architectural standards, and a consistent approach. To genuinely replace the comprehensive features of modern SaaS enterprise resource planning (ERP) solutions, significant development and rigorous enforcement of skills within the platform are needed. Even then, product manager review, user acceptance testing, and architectural inspection are needed, and it appears that the time required for such activities is greater with vibe-coded solutions.

The Context: IBRS Fourth-Wave ICT Architecture

IBRS is not arguing that vibe AI code generation has no place. It is a pivotal technology. It will enable far faster, more robust development of niche applications that would never have been economically viable with traditional coding methods.

To understand where this actually fits, we must refer to the IBRS Fourth-Wave ICT architecture. As IBRS has been discussing since 2020, the move to a Cloud-based, As-a-Service ICT and business operating model. Their framework explicitly calls out the use of large core business platforms, supported by specialised industry applications, and finally surrounded by a universe of small, custom applications that perform the unique processes that drive competitive value. It is these unique satellite applications where vibe coding will be the star.

Vibe coding is not a replacement for SaaS; it is a new capability for rapidly developing and deploying extensions at the edge of the Fourth Wave digital fabric. In this architecture, core SaaS platforms (like Salesforce or ServiceNow) provide the stable System of Record and the Digital Fabric (governed integration and security). Vibe coding and agentic AI, along with other low-code and orchestration services, serve as the System of Innovation, enabling teams to build hyper-niche, short-lived tools on that stable foundation.

The FUD (Fear, Uncertainty, and Doubt) being spread is that the innovation layer will eat the record layer. It won’t. One provides niche solutions (that will arguably have a huge positive impact on the organisation’s operations). The other provides the enterprise’s truth and structure.

Next Steps

• Architecture Review: Conduct a high-level architecture review, mapping your organisation’s software solutions to the IBRS Fourth-Wave ICT architecture. Identify the central core platforms and the supporting ring of industry-specific solutions, then explore the many smaller, often bespoke solutions and low-code workflows.
• What are the Solutions to Build? Using the above review as a basis for conversation, explore where AI-generated code can be used safely and with strict governance to further extend the business environment with new niche solutions.
• What are Our Development Resources? Compare the costs, skills availability, and governance required to create these new niche solutions using vibe coding vs. low-code tools or the orchestration and agentic tools readily available within core SaaS platforms.
• Create a Decision Framework: Create a formal decision model to determine when to use embedded SaaS platforms orchestration or agentic AI services, when to use low-code, and when to build custom solutions with vibe coding.

Footnotes

1. ‘Why are Software Stocks Down?’, deVere Group, 2026.
2. ‘Markets recalibrate amid shifting crrents’, BIS, 2026.
3. ‘ServiceNow Reports Fourth Quarter and Full-Year 2025 Financial Results’, ServiceNow, 2026.
4. ‘Salesforce Delivers Record Fourth Quarter Fiscal 2026 Results’, Salesforce, 2026.
5. ‘85 SaaS Statistics, Trends and Benchmarks for 2026’, Vena Solutions, 2026.