Every few years, technology observers proclaim the death of a major paradigm. "Email is dead," they'd say, before watching usage climb to new heights. "Personal computers are finished," just before we all bought new laptops for remote work. These pronouncements often miss the subtle but profound transformations happening beneath the surface – not death, but evolution. Today, we stand at such a juncture with Software as a Service (SaaS), witnessing both its apparent demise and its powerful reinvention.

For over two decades, SaaS has dominated the software landscape. The model was revolutionary in its time – moving applications from on-premises installations to cloud-hosted solutions with subscription pricing. This shift democratized access to powerful software and created enormous value for businesses and consumers alike. But the traditional SaaS paradigm – web and mobile apps backed by databases – emerged from an era when computers needed precise instructions to function. The interfaces we've grown accustomed to were artifacts of machines' limitations, not necessarily the ideal way for humans to interact with digital systems.

The rise of generative AI and natural language understanding is now fundamentally altering this landscape. The apparent "death" of SaaS comes as we move beyond the need for fixed-function software with rigid interfaces when we can simply describe what we want. Consider the promise of what I've come to call "vibe coding" – the emerging ability for non-technical users to generate custom software solutions on demand through conversation. "I need an app that tracks my team's project milestones and sends automatic reminders" could soon be enough to manifest a working solution without traditional development cycles or SaaS subscriptions.

This democratization of software creation directly challenges the fundamental SaaS value proposition of "we built this software so you don't have to." If anyone can create customized applications through natural language, why pay for pre-packaged solutions? The legacy SaaS model built on specific, unchanging interfaces seems increasingly outdated in a world where anyone might become their own software developer through conversation with AI.

Yet SaaS was never just about software – it's about connected systems with shared data and unified perspectives. This is where the "long live SaaS" part of our story emerges. While individuals might generate isolated applications tailored to their specific needs, the network effects of collaborative platforms cannot be easily replicated. Consider what makes tools like Notion or Salesforce valuable – it's not just their features, but the shared, structured data environments they create where multiple users collaborate within consistent frameworks.

These platforms create value through shared data lakes, governance structures, permission models, and embodied domain expertise that personal AI-generated applications lack. The future isn't the elimination of SaaS but its evolution – merging traditional structured software with AI flexibility to create platforms that maintain critical shared infrastructure while allowing unprecedented personalization through natural language interfaces.

The "one-size-fits-all" interface paradigm is another casualty in this transformation. Traditional SaaS products have long struggled with a famous paradox: 90% of Excel users only use 10% of its features, yet those feature sets differ dramatically across user segments. This created an inevitable tension between simplicity and power, forcing compromises that satisfied no one completely.

With AI-powered personalization, we finally break free from this constraint. Instead of building interfaces that inadequately serve everyone, intelligent agents can now curate personalized experiences that evolve as users progress from novices to power users, revealing complexity only when needed and in contexts where it makes sense. Interfaces can adapt to individual workflows, learning from usage patterns to present the most relevant tools at the right moment.

This represents a profound shift for UI/UX designers. Their focus moves from creating static, comprehensive interfaces to designing cohesive systems that can be dynamically assembled and presented. The art becomes one of choreographing the user's journey through increasing capability – designing not just screens but growth paths that adapt to each user's unique needs and learning curve. This eliminates the forced choice between overwhelming complexity and limiting simplicity that has defined software design for decades.

Perhaps most exciting is how this evolution opens territories previously inaccessible to software. The combination of natural language processing, AI agents, and flexible interfaces has dramatically lowered the barriers to serving niche markets and handling unstructured data. Markets that were once too small to justify dedicated software development can now be served profitably. The economics fundamentally change when you can deploy AI agents that adapt to specific needs without custom coding for each variation.

This shift even challenges our conception of what an interface should be. Some innovative solutions are abandoning traditional interfaces entirely – a friend recently described a startup serving customers purely through email, with AI agents handling the complexity behind the scenes while a small web app exists only for administrative functions. The customers never have to learn a new interface; they simply communicate naturally through a medium they already understand.

Other services might become ambient, operating largely autonomously while surfacing only critical decisions or insights that require human input. The SaaS of tomorrow might look nothing like the dashboards and portals we associate with software today, yet will deliver even greater value by adapting to how people naturally work rather than forcing humans to adapt to software limitations.

As we navigate this transition, the enduring SaaS platforms will be those that embrace this new reality – providing the critical infrastructure, connectivity, and trust layers that enable collaboration at scale, while leveraging AI to handle the last-mile customization that makes their services relevant to each user's unique context. This isn't the death of SaaS but its rebirth – evolving from rigid software products to flexible collaborative platforms that combine the best of centralized infrastructure with personalized, AI-powered experiences.

SaaS is dead. Long live SaaS.


Now on to the jobs. Fewer jobs now as I have closed a number of positions (and another number closed without my help :-) and as I wrap a wonderful first financial year of operations, things will be a little slow till the start of April. Which means that I am accepting new clients. If your company is hiring, please reach out to me, ideally by signing up at https://recruit.svs.io and uploading your JDs. I'll be in touch shortly thereafter.

Also, there's one role that I'm happy to recieve some help closing. Sundial is one of my favourite clients - I've had two people join them and they're both very happy. Sundial is looking for a lead software engineer to build out their AI/MLOps stack. I am of course happy to share the commission with whoever finds me the right candidate. If you have an invite link, please use it to invite your friends and as always, feel free to forward this email to whoever might benefit from it.