The “Death of SaaS” = Death of FTE Industrial Economics

Author, Aya Shimada, on a swing near Dong Van in the Hagiag region in northern Vietnam Jan 2026

• Thanks to AI, software pricing, and enterprise IT's purpose, are shifting from counting FTEs—an industrial relic—to capturing value: labor inputs → outcome multiples.

• Human workforce transitions from software operators to value orchestrators/strategists—or risk obsolescence.

---

Given Fintech Week, I've been thinking a lot about the so-called “Death of SaaS and how it relates to industry shifts (what each company should do) and human labor (the quality of our work), especially from a Japan-based perspective. (For context, I used to be global head of commercial strategy at a fintech, with hands-on coding roots in banking IT in New York. Today, I advise companies on digital transformation and navigating the AI/software currents.)  　

Here’s my thesis. Thanks to AI, software pricing, and enterprise IT's purpose, are shifting from counting FTEs —an industrial relic—to capturing value.

Human workforce transitions from software operators to value orchestrators/strategists—or risk obsolescence.

https://video.wixstatic.com/video/b06bbd_52199958516445469e92cbb4c1cb8e57/720p/mp4/file.mp4

Surviving the Death of SaaS, this article as interpreted by NotebookLM's Video Overview

What the “Death of SaaS” Really Means

The “Death of SaaS” is not about SaaS disappearing overnight. It is about the erosion of the classic SaaS model:

• AI agents are undermining per‑user licensing by making usage and outcomes more measurable and automatable.

• AI‑assisted coding can replicate much of what SaaS products do, faster and cheaper than before.        

In the near future, humans will no longer interface software. AI will. If needed, a human software interface (“UI”) will be created by AI on the fly.

The question is not “Will SaaS die?” but: Who will survive as software economic model is re‑written and what will happen to people and organizations?

How Each Player Can Win

Legacy SaaS Vendors

From the competitive perspective, legacy SaaS players still have highly specialized assets and operations, and customer trust—but they cannot rely on per‑user SaaS pricing forever.

They can win by:

• Building agentic AI directly into their products, so that customers buy outcomes and automation, not just screens and workflows.

• Transitioning from per‑seat pricing to usage‑ or outcome‑based models, aligning revenue with the value that AI agents actually deliver. This pricing evolution mirrors the broader economic shift from FTE counting to value capture.

• Shifting to AI-assisted coding for efficiency – fewer people will be required per task, and shipping agility is achieved.

Those who cannot make this shift risk becoming disrupted by another vendor’s AI agent layer and AI-assisted startups.

Startups

Startups are structurally better positioned to embrace the “Death of SaaS” era—but only if they avoid becoming a thin UI layer over commoditized models.

They can win by:

• Using speed (with AI-assisted coding), creating vertically specialized agents that understand specific industries, regulations, and workflows instead of generic UI layer.

• Building proprietary data moats that foundation models cannot easily replicate - domain-specific, non-public, closed-loop, and portable across underlying model changes.

• Leaning into outcome‑based pricing.

Big Systems Integrators and Legacy Software Vendors (Especially in Japan)

For large systems integrators and long‑standing software vendors, the old era of vendor lock‑in is fading fast.

ERP and SaaS customization can now be replicated with AI‑assisted development, and even COBOL or “spaghetti” legacy architectures are becoming candidates for automated re‑engineering. Vendor power that relied on opacity and complexity is diminishing.

They can win by:

• Pitching modernization as a defensive move against legacy contract cancellation.

• Partnering with or acquiring differentiated startups and former subcontractors that bring AI‑native capabilities, rather than trying to build everything internally with legacy mindsets.

• Retraining staff in AI‑assisted coding so they can compete with faster, leaner teams instead of being displaced by them.

Those who remain tied to old “body‑shop” models will struggle in a world where AI accelerates development and reduces headcount needs.

Enterprises: Cutting Costs and Driving Innovation

Enterprises can use the “Death of SaaS” to both reduce costs and accelerate digital innovation.

1. Cutting Costs

• Replace SaaS tools with AI agents that plug into your current tech stack.

• Replace legacy software vendors with faster, cheaper startups or new entrants empowered by AI‑assisted coding and more flexible operating models.

In Japan specifically, there is a major opportunity for IT organizations to move from an old FTE‑based cost‑center model to business value‑based budgeting, where investment is tied to measurable business outcomes rather than headcount.

This embodies the core shift: from counting FTEs (industrial relic) to capturing value through measurable outcomes.

2. Driving Digital Innovation

Cost optimization alone is not enough. The real leverage is in building the capability to innovate continuously.

Two practical levers:

• Vibe Coding for democratized innovation

  Instead of relying solely on PowerPoint, meetings, and expensive POCs from external vendors, existing IT and business employees can learn security‑ring‑fenced “vibe coding”—a lightweight, experiment‑driven approach to building small tools and prototypes.

  
  

  This enables a “make‑to‑think” culture where teams express ideas in working software, then iterate based on real usage.

  
  

• AI‑Assisted Coding for real enterprise builds

  Small, in‑house dev teams using AI‑assisted coding can deliver meaningful enterprise systems with far fewer FTEs.

  
  

  This may partially replace traditional vendors that customize and maintain CRM or ERP to the company’s specific workflows — agility, for innovating new processes with speed, will finally be possible. At the same time the economics change and heavy customization vendors become unnecessary.

  
  

In many legacy Japanese enterprises, transitioning to in‑house AI‑assisted development is difficult because the current IT workforce often does not have deep coding backgrounds. For those organizations, options include acquiring software vendors or hiring rare leaders who can build AI‑native engineering teams from scratch.

The Real “Death” Is About Economics, Not Software

This 'Death of SaaS' is fundamentally economic: software pricing and IT's purpose shifting from counting FTEs to capturing value—labor inputs to outcome multiples.

The “Death of SaaS” is not a funeral for software delivered over the internet. It is the slow death of a particular economic and organizational model:

• Per‑user licenses detached from measurable outcomes

• Heavy reliance on customization vendors

• IT as a cost center measured in headcount, not value

Those who adapt—by embracing agentic AI, outcome‑based pricing, proprietary data, in‑house AI‑assisted development, and democratized “vibe coding”—stand to win in the next era.

Those who don’t may still have customers and products for a while, but the economics will move against them.

Did this resonate with your research or experience? What insights or perspectives would you add? I'd love to hear your thoughts.

I plan to continue exploring digital landscape shifts and how corporations must adapt.

Please feel free to contact me at aya.shimada@culturelabs.co.