AI isn't killing SaaS – it's killing single-purpose SaaS | Hacker News Reader

AI isn't killing SaaS – it's killing single-purpose SaaS

9 pointsJB_50007 comments

Over the last year I keep seeing “SaaS is dead” takes. I don’t think that’s what’s happening.

What seems under pressure isn’t SaaS as a model. It’s narrow SaaS built around a single capability that AI can now reproduce cheaply.

If your product is basically a thin wrapper over a model, or differentiated mainly by features rather than workflow integration, the moat feels weaker now. AI compresses build time dramatically. That means more competitors, faster cloning, and lower switching costs.

But SaaS that is model-agnostic, deeply embedded into workflows, or acts as connective tissue between systems looks much more durable. Integration, distribution, and trust don’t commoditize as quickly as features do.

It feels less like SaaS collapsing and more like a sorting event. Thin wrappers get squeezed. Infrastructure layers and integrators get stronger.

Curious if others building right now are seeing the same shift, or if I’m over-indexing on AI-native workflows.

Yeah, we've had a few super early stage startups who've straight up asked as to why can't we just vibe code this. So that's there.

But I truly resonate with the workflow approach, how I see the new SaaS model is like pieces of Lego. You can plug and play and then the world will start making sense ig?

I feel for the most part there's a middle ground where SaaS is profitable.

Micro companies won't buy SaaS, $20 is a lot for them. At the top, it's also too big or too important to outsource. I had a sales call with a hospital chain once that didn't want to use Slack or MS Teams because it can't meet their needs. They were building their own solution despite not being a tech company.

The happy zone is the middle range where paying $5/head is nothing. These are the companies making $10m/year and such.

But the idea is AI extends the non-paying markets at the top and bottom. The companies at the bottom can go on longer without paying for Figma. The companies at the top are more likely to build their own solutions.

I've wondering about this too and tend to agree. Vibe coding a single function app is easy, but delivering a platform that scales to enterprise levels is much different, especially with proper security, QA, and UAT. Not saying AI won't be capable of this fully, but it's certainly harder. Plus, do companies really want to take on the burden of upgrading, debugging, and maintaining everything themselves, I think not. But what do I know, I'm just a human hoping he'll still have a job :p.

I'm not sure I'm buying this.

Software is a packaged experience, business logic that gets refined over time. You pay for the likelihood that the logic is correct, a server that someone else manages, and hopefully someone who will answer your emails if something is wrong.

Sure, you can knock some things together quickly, but you still need to test your code against reality and your customers' expectations. Getting the recipe right takes a while.

So many comments here are AI-generated, including this post.

The thing is that most people who claim that AI will kill SAAS have never run a SAAS business themselves or have never properly used a SAAS as a real customer for multiple years.

SAAS companies are not just about the software. They provide a series of things including hosting/maintenance, support, data, integrations and most importantly: solving a problem that customer is not interested in solving themselves because they hae more core problems to solve for their own business.

Is it true that some SAAS companies need to go out of business or do better than what they do now (e.g. locked in contracts, shitty support etc) ? Yes, absolutely. However, AI or vibe coding is not magically going to solve this problem.

At the end of the day, SAAS is a business model that has extreme value.

Source: I may be biased as a SAAS founder but I can assure that majority of our customers are not waking up daily and thinking "How can I stop paying this company $xx,xxx/Year for my business where it makes me $xxx,xxx/Year without me needing to hire/build my own product/team". Most of them certainly don't even think about "lets vibe code this shit". How do I know ? They are telling me on a daily basis. Sure, they need us to get on the AI bandwagon to solve their problems a little faster than now. If we cannot keep up, someday they may leave who can solve their problem faster/cheaper but they are not going to vibe code their own homegrown solution. In fact, we routinely migrate B2B customers from their own homegrown solutions (even before AI).

We're moving from the 'there's an app for that' era to the 'there's a prompt for that' era. The issue for single-purpose SaaS isn't just the AI features, it's the high 'context switching' cost. If I can achieve the same result within my primary IDE or communication tool using a specialized agent, why would I maintain a separate subscription and login? The survival of SaaS in 2026 depends on deep ecosystem integration, not just isolated utility.

Reply on news.ycombinator.com