AI Isn’t Killing SaaS. It’s Forcing It to Grow Up.

AI is real—and transformative—but the idea that it will replace SaaS overnight ignores enterprise reality.

Having led large-scale transformations, I see a different future: SaaS evolving into something akin to agentic AI as a service.

AI Hype vs. Enterprise Reality

Everywhere you look right now, it’s the same narrative: AI is here, SaaS is dead. As someone who has spent decades building and scaling platforms like Salesforce across enterprises, I’ll admit—it’s a bit exhausting.

Let’s be clear: AI is real. Very real. It’s already here, and it’s evolving at breakneck speed. There’s no debate about that. Personally, I’m genuinely excited. I’ve been deep in learning, experimenting, and pushing into this space—it’s been a long time since something in technology has felt this transformative.

But we’re still early. There are real questions that remain—around safety, ethics, cost, and even basic suitability for many use cases. This audience knows that.

So while the hype cycle is in overdrive, the reality is still being shaped.

Transitioning to Agentic AI Will Take Time

Now, let’s entertain the extreme narrative for a moment: that AI—agentic or otherwise—can replace every enterprise system we have today. It won’t. But let’s pretend.

Even then, what does that actually look like?

I’ve led multiple migrations—from legacy platforms into SaaS ecosystems when SaaS itself was the disruptor. Even under ideal conditions, those transitions take months, often years. And that’s just the technology.

Layer in change management—and the timeline stretches even further.

Now imagine doing that in a world where AI promises leaner operations. Fewer people, more tools, higher expectations.

Adoption challenges stretch the timeline even more.

And while all of this is happening, what do you think best-in-class SaaS companies are doing? Sitting still? Not even close.

In the Meantime, SaaS Isn’t Dying—It’s Evolving

Take Salesforce as a representative example. It’s not resisting AI—it’s leaning in hard. Platforms like Agentforce are already enabling customers to experiment with agentic AI in practical, incremental ways - alongside their SaaS investment.

What’s especially compelling is the blend of deterministic constructs (like flows) with generative AI.
That balance matters more than most people realize in enterprise environments.

We’re already seeing this expand across functional domains like marketing and sales, industries like financial services, and beyond. There’s coverage for low-code, pro-code, RAG, interoperability, and even a broader vision where tools like Slack become the conversational fabric connecting agents, humans, and workflows. As of this writing, Salesforce announced Headless 360 - an entire platform for agentic AI

This isn’t SaaS dying. This is SaaS evolving.

If we’re allowed to believe the “AI replaces everything” narrative, it’s only fair to believe SaaS platforms will evolve just as aggressively.

What does that evolution look like?

• Out-of-the-box agents for functional domains (horizontals) - sales, service, marketing etc

• Industry-specific (verticals) agents built on years of domain expertise

• Customizable platforms to build your own custom agents

• A conversational layer (the fabric) that complements—not replaces—the UI

• Interoperability across ecosystems (MCP et al)

True Cost of AI and ROI is Unclear

Now, let’s talk about something that gets far less attention: cost.

AI pricing models are still evolving. Most are consumption-based, which sounds great—until you try to forecast or control it at scale.

Many organizations are already deploying AI into use cases where the ROI is unclear—simply because they can.

Then there are the broader costs:

• Environmental — energy and water usage

• Ethical — bias and accountability

• Social — job displacement

• Economic — economies are predicated on a threshold of unemployment, what happens when that threshold is breached?

These costs are arguably yet to be factored in to the equation. If we did, ROI is arguably lower or even non-existent for the simpler use cases. The low hanging fruit we may be currently deploying AI for.

This is where Saas platforms like Salesforce have an important role to play.

With their scale, customer base, decades of domain & industry knowledge, and GTM muscle, Saas platforms like Salesforce can standardize and distribute common AI use cases—amortizing cost across thousands of customers, much like they did during the SaaS wave.

That’s how costs come down. That’s how adoption becomes sustainable. Not by everyone doing the same agents for the same/ similar use cases.

Mitigate socio-economic costs for Now

Stepping back—AI is absolutely a tectonic shift. It’s here to stay. But where we choose to apply it matters.

At some point, we’ll hit diminishing returns.
Not every problem needs AI. And not every AI solution makes financial sense.

We should be using it to solve truly hard, high-impact problems—accelerating drug discovery, transforming healthcare outcomes, tackling systemic inefficiencies.

Not just automating the simplest tasks and calling it progress.

And even where we do apply AI to simpler problems, there’s a choice:

Do more with the same people—or do the same with fewer people.

Right now, given how early we are—and how many unknowns still exist—I’d argue for the former. That would at least let us as a society, mitigate some of the social and economic costs while things evolve.

If you’re looking to be thoughtful about your SaaS/ Salesforce investment and find the AI noise overwhelming—confusing, overhyped, or just impractical—I’d welcome a conversation.

No hype. Just a grounded discussion on how to thoughtfully integrate AI into what you already have.

At the very least, I’d love to hear what’s on your mind—and help where I can.