AI生成应用时代，SaaS的护城河还剩哪些？

What I'm going to talk about today is the ra the rise of AI generated apps and what it means for SAS. And I know there's quite a few people who are here who are asking themselves the same question. You have a SAS whether it's small, whether it's mediumsiz, whether it's already pretty large and you're like what is going to be the impact of the basically generated apps on what we do. So first a little bit about who I am.

Uh as was mentioned, I'm the creator of DN Simple. I've been running that now for 16 years. Uh I'm also a software developer and it turns out it's actually a lot longer than that. So, I don't even want to talk about that, but I do want to talk about why you might want to trust me. So, as you can see here, I have a business that I've been running for quite a long time. I'm the CEO of that. Uh, we've had competitors from all over the place, from the biggest companies like Google, Microsoft, and Amazon, all the way down to the smallest new competitors that are trying to do things to eat our lunch.

I'm also starting another business at the same time. So, I'm doing this from the other side as an independent who's also trying to use generative AI to do things. So, I am both the enemy and the friend. So why you shouldn't trust me though is because I don't have a crystal ball. I cannot see the future. I have no idea what these companies are going to do. All right? So they could and as we've seen the pace has been incredible how fast this stuff has been happening.

So what I tell you now might or might not work in the near future, but I have some fairly high confidence. So first let's talk about state-of-the-art. So fact generative AI makes building software easier than ever before. Coding assistants are helping non-engineers build software and dark factories are starting to build software with little human involvement. That's happening right now and it's happening at a breakneck speed.

So that's the state-of-the-art. It's happening. It's coming. It's already out there. You have to be ready for it. A lot of the previous SAS modes that would have worked before aren't going to work in the same way. But I want to take a look at them real quick. I have four SAS modes that some of you or all of you have probably heard about. Uh the first is network effects. So that means that the more people that can use your software together and connect together, the more value they get out of it.

The second is proprietary data and knowledge. That means the data and knowledge that you have inside your systems is where the value is. The third is deep workflow integration. So that's your software works with other software and other systems to be able to make things work smoother and better for your customers. And finally, one of the previous SAS modes was high switching cost. Once you've invested time to put your data and energy into a SAS, it's kind of hard to leave.

And a lot of SAS has capitalize on that to keep you from leaving. But not all those are defensible in this new environment. So for example, features used to be something that would set SAS apart. They would build features that nobody else had. Uh they would do the high switching costs and they'd have the integration. Those aren't defensible anymore, especially the features side because it takes less and less time to put out features for software.

So that's why you're you're if you haven't seen it yet or if you haven't experienced yet inside your SAS, you will probably start building and releasing features faster than you have before once you start adopting the generative AI tools for software development. But high switching costs are no longer there because you can use those same tools to help guide you and even execute leaving one provider to go to another provider.

Same with integrations. So this one I'm a little sketchy on. I'm not 100% sure because integrations are still really difficult because change is happening at a break neck speed also at the places where you integrate with. So now you have to deal with that as well. But generally I don't think those are defensible anymore. So what modes are still viable? So the first one that I want to talk about is is uh is collaboration.

So this is one of the areas where you should probably focus. So software where value grows as human collaboration or agent collaboration grows is where there is more value. Why is that? Because it's that network effect. The network effect is still real. Interconnected systems are something that that somebody that's generating an app for their own use or inside of their business. They're not looking for network effects.

They they don't need network effects because they're trying to solve one single problem. But as a SAS, you do need to look for network effects in places where collaboration helps. The second thing I want to talk about is regulated environments. So I I've spoken to a lot of SAS founders over the years here at Microcom and I know a lot of you are niched into regulated environment areas. Whether those are things that require HIPPA compliance, whether it's things that are require very specific knowledge about some industry around safety, uh things like that.

That's actually a really good space to continue working in. Why? because generated apps are going to be really hard to live in these regulated environments because they're going to be done in a way that's sort of chaotic and random and regulation and randomness do not go well together. So this is the second area where I'd say if you're building a SAS and you're looking for a niche, keep that keep your mind open to that.

Hey, Rob Walling here. If you're watching this and thinking, I wish I could be in the room for talks like this, you can and you should. So much of the magic at our events happen in the hallway track. That's where you connect with two to 300 like-minded founders who are trying to solve many of the same problems you are. I hope you'll join us at our next event. Head to micromp.com/events to grab your ticket. Our next one is in Iceland in September of 2026 and then we'll be in Austin in April of 2027.

The third thing I want to talk about which isn't really an environment is operational excellence. Uh so when I talk about this I talk a lot about it from the from the engineering side of operational excellence. As more code is produced by generative AI the quality of that code will necessarily go down. It's inevitable. There will not be as highquality code. So more code will be produced. Pressure is going to shift from the development of the code to the operations of those codes.

So the better that you can be inside of your organization with operational excellence, especially technical excellence, the better chance you'll have inside of your SAS company to survive because frankly, people are going to be using these one-off apps and a lot of them are going to suck really, really bad. They're going to do the job, but they're going to be worse and they're going to break more often than anything that they've ever used before.

So they're going to be looking for cases where operational excellence is still in place. Another thing that I want to talk about and this is the final one is I want to talk about internet infrastructure. So um capex is still a mode. So that's capital expenditures. It's still if you if it costs money to do what you do. So in other words money that not everybody else has to invest. You have a moat still. So I'll take DN simple as an example.

We work in a world where DNS has a very specific design that requires servers all over the world. That's a capex expense that we have to to do in order to make it work at the scale that we work at and even at a small scale because your customers are any anywhere in the world. It's a it's basically a factor that restricts people from entering into our space. So the areas of internet infrastructure which is going to be necessary and that's hosting uh I've met folks here who do sma