What The SaaSpocolypse Doomsdayers Overloook

Investors believe SaaS is dead and AI is the new king. But there's a more fundamental question that's being overlooked: how much software ownership do customers really want to take on themselves?

This is Part IV in a series about how AI is driving category-level changes in the software industry. Head here to start on Part I.

---

Everyone is saying that SaaS is dead. Or at least it will be, once AI is done eating it for lunch.

But I think this “SaaSpocolypse” view misses the real question.

It’s not about whether ”AI-first” or “traditional SaaS” will win out. There is a more fundamental question at play: how much software ownership do customers want to take on for themselves?

AI-first software, like that from 8090 Solutions, is special because it changes something fundamental: it shifts the responsibility of owning software functionality from the vendor to the customer.

This is the shift I want to explore.

While AI-first software makes lots of promises, I’m not convinced that every business will want to go all in.

AI-first software changes something fundamental: it shifts the responsibility of owning software functionality from the vendor to the customer.

To show you why, we’ll look at three supposed advantages of AI-first software: lower cost of ownership, more flexibility, and greater capabilities. Then, we’ll answer two questions: Are these advantages as large as they seem, and are they durable? And are those advantages things customers will want, given the tradeoffs?

For the purposes of this article, we’ll contrast two approaches to software:

1. AI-first software is designed to give users control over what the software does. For example, 8090 Solutions lets customers specify product requirements and then builds software accordingly. It’s the customer who owns product development.

2. Traditional SaaS is designed to maintain vendor control over the software. The vendor determines which new functionality to develop and then releases it to the customer. Sometimes, customers can create their own workflows or automations, but they cannot directly alter the software itself.

Whether a customer wants ownership, though, is a question that’s worth scrutinizing. While the investor community has already made up its mind, I think this is a rush to judgment.

AI-First Software Makes Three Promises Over Traditional SaaS

The benefit of AI-first software comes down to three things:

• The first is the cost of ownership. While a traditional SaaS business might need hundreds of engineers to maintain and update a product, an AI-first business can provide the infrastructure for a customer to create its own software with far fewer employees.

• The second is development flexibility. Traditional SaaS requires customers to accept the features and functionality provided by the vendor, but AI-first software enables customers to design what they need, quickly.

• The third concerns capabilities. AI already offers capabilities that traditional SaaS hasn’t provided yet, like agentic AI. This gives AI-first products a head start on the value they can deliver.

But which of these benefits are durable (they provide an advantage that SaaS companies could not obtain for themselves) and which are desirable (customers actually want them, given any tradeoffs they might impose)?

To find out, let’s explore each:

Cost of Ownership

The premise behind this advantage is simple: if AI-first vendors can develop software with a smaller engineering and product team, they have a lower cost basis and can profitably sell their software for less. But there are two reasons why this equation isn’t so simple.

The first is that we don’t yet know the true cost of running an AI-first business at scale. While AI-first software makes it easier to develop user experiences, it requires a much more robust architecture underneath. What is the true cost of supporting that, especially when you consider the computing costs that AI requires? I’m not saying that AI-first companies won’t have a cost advantage, but I am saying it’s too soon to be certain of how large this advantage is.

The second, and more important, consideration is that engineering and product staff do not make up the majority of a traditional SaaS company’s expenses. Sales and marketing do. This used to be the other way around. But as SaaS software categories became immensely crowded in the 2010s and 2020s, traditional SaaS companies have had to ramp up sales and marketing spend dramatically to remain competitive. AI-first companies don’t face such pressure (yet) because this category is relatively nascent. But what happens to sales and marketing costs when more competitors emerge?

Development Flexibility

Personally, I like the promise of flexibility that AI-first software makes. Who wants to wait for a feature to come out when you can just vibe code it yourself? But there’s the rub. Deciding what you need and how it should work is not a trivial undertaking. As you move to the enterprise, that “ownership tax” can become a real burden.

Microsoft Excel is actually a good microcosm of this. If you’re building a financial model, Excel’s built-in formulas make the cost of manipulating data almost zero. The real effort comes from deciding what the financial model should be. That’s why even though it’s easy enough to learn Excel, CFOs are as valuable as ever.

I think there’s a similar dynamic at play with software. Just because a business can take ownership of its software, does that mean it wants to? I’m not sure this is a given. It’s the same kind of “make vs buy” decisions businesses have always faced. Product ownership requires a competency that many businesses may lack, and it may be cheaper to delegate this to someone else.

Capabilities

This is where AI-first software may have the strongest advantage. It has a fundamentally different architecture that allows it to do things that traditional SaaS cannot.

Agentic AI is one such example. As we discussed earlier, agentic AI needs a unified data layer to give it context about the business. AI-first companies built their products with such an architecture in mind, but traditional SaaS companies did not. When data is fragmented, siloed, and inconsistent, agentic AI is a non-starter.

If every SaaS company were built this way, it would indeed prove fatal for the category at large. But here’s the thing: the software ecosystem isn’t so black and white. Some forward-thinking SaaS companies have already re-architected their products to be AI-native. This isn’t conjecture; I’ve met product leaders who have already done this.

But because this has happened behind the scenes, it’s easy to lump in these progressive SaaS companies with the laggards. They may look the same on the outside, but underneath, they’re very different. And they’re not at the same disadvantage as you might be led to believe. With the right foundation in place, these “SaaS + AI-native” businesses might be in a position to win after all.

This Chapter Is Really About How Much Ownership Customers Want

This is what I think the market is overlooking.

Yes, AI-first software can do more (at least today) than traditional SaaS can. But it’s changing something more fundamental: who takes responsibility for what the software does, how it works, and how it’s supposed to achieve outcomes.

Some organizations will love this. But for others, it will feel like an ownership tax they’d rather pay someone else to handle, even if there are some cost advantages. I know my way around Excel, but I still hire an accountant.

That’s why I don’t see this as a battle between AI-first and traditional SaaS.

I see this as a new chapter in which businesses figure out how much ownership of their software they want to take on themselves, and how much they want to delegate to a vendor.

The answer, in my view, is not going to be at either extreme. Which approach emerges as the dominant one, and whether it’s come from progressive, AI-ready SaaS companies or new entrants in the AI-first camp, is very much up for grabs.

What does each camp need to do in order to win?

That’s what we’ll unpack next week. See you then.

---

As the founder of Flag & Frontier, John Rougeux partners with executive teams to align on their strategic narrative, build belief in the market, and win the next chapter of their business. You can chat with John here or connect with him on LinkedIn.