Episode 842 | What is the Future of SaaS in an AI World? (Rob Solo)

Rob Walling: It’s another episode of Startups for the Rest of Us. I’m your host, Rob Walling. And in this episode, I talk about the future of SaaS in an AI world. And this topic has been floating around the internet as people are talking about SaaS being dead or asking if SaaS is going to exist in five years. And I’ve actually received multiple listener questions and I’ll read one of those in a second that’s asking this exact question. Before we dive in, I have a couple quick MicroConf announcements. First, Early Bird tickets for our next MicroConf are officially on sale. We’re heading to Austin, Texas, April 18th through the 20th of 2027 for three days of world-class talks, meaningful conversations with other founders, and some great excursions around the city. If you’re thinking about joining us, now’s the time to grab your ticket. Early Bird is the lowest priced ticket tier we’ll offer and the prices will only go up as we get closer to the event. 

Plus, producer Sonya gave me the promo code Rob50, that’s Rob50, which gives listeners of this podcast an additional discount at checkout. MicroConf events consistently sell out and I expect Austin to be no different. Head to microconf.com/us and reserve your spot. Also, I want to let you know that if you’re a MicroConf Connect member, we’re going live next week with Harris Kenny, who you may remember from TinySeed Tales season five. Tracy Osborn is sitting down with Harris for a fireside chat and they’re going to talk about mindset and bias toward action, what that actually looks like day to day, how to know when to push through versus change course and the small habits that can pull you out of a funk. It’s the kind of conversation that you will not find anywhere else. And if you’re not a member, we’d love to have you. It’s at microconfconnect.com. 

MicroConf Connect is the year-round version of our hallway track from our in-person events. It’s our vetted online community of founders having the kinds of conversations you can’t have with just anyone. These are the wins, the struggles, the, quote, “Am I the only one dealing with this?” moments. And live sessions like this one are just one of the perks. If you’re interested, microconfconnect.com to apply. 

Let’s dive into the future of SaaS in an AI world. This question was a simple reply to my robwalling.com email. So if you go to robwalling.com/emails, you can sign up to receive a weekly email from me. And a reader responded and said, “What’s the future of SaaS companies now that AI makes it easier than ever to build and copy a product?” And this, as I said at the top, is a variation of questions I’m seeing all over the internet. And I have received other versions of this on this show. And so I really wanted to sit down today and try to give as much of a definitive answer as I can to this question. So what is the future of SaaS now that AI makes it easy to build and copy SaaS products? I’m going to start by giving a short answer in four words. 

The future of SaaS is SaaS. And now I want to show you why. But before I do that, one rule for how to listen to anyone on this topic, including me, is to be very careful with anyone who tells you something is dead because we have heard this song before. In 2000 after the dot-com crash, the media told us the internet was done, over. Nothing left to build. They said this on the internet. I don’t know if you weren’t alive then or weren’t paying attention, it was so pervasive and just a given that the internet was never going to be worth anything. And then mobile and iOS were going to kill the web in, I don’t remember, ’08, ’09. It was going away. That didn’t happen. Then blockchain and Web3 were going to kill everything. I lost track of exactly what it was going to kill, but it didn’t happen. 

Maybe the web, I don’t even remember. No code was going to kill software development. Everybody’s going to build their own apps. Nobody needs engineers. It didn’t happen. There are other examples. And now AI, the new technology is going to kill SaaS. So it’s the same song. It’s a new verse. Here’s the thing I see about people saying it. Most of them, not all of them, but I would say the vast majority, 80, 90 plus percent of them have never built a SaaS company. And actually, I wouldn’t be surprised if 95% of them have never built any company. Saying X is dead is a way to get free clicks. It’s a way to get eyeballs for declaring something outlandish and nobody comes back to check if you are right or wrong. So just be very careful. If you hear this message every day from some new chucklehead on the internet, they’re still a chucklehead. 

Just because a lot of people are saying it doesn’t mean that they’re right. Calibrate accordingly. It’s social media. Pick who you’re going to listen to. Usually things are not dying or dead, but to be fair, there’s a real version of this argument and I break it down into four separate claims that I’m going to talk through individually. The first one is that everyone is going to self-host their own custom app that they either vibe code or build with AI. The idea here is that in addition to, let’s say you’re a SaaS company and you pay for SaaS subscriptions, that you’re going to vibe code or code all of those apps, custom host them. And maybe that the dentist down the street is also going to build their own practice management software and self-host it. So that’s kind of argument number one. Point number two is that any app can be replicated in a day. 

I’ve heard people say in a week, in a month, in some short timeframe so that there’s no moat and that copying an app is free. The third claim is that agents are going to replace everything, that they’re going to do all the work and so we won’t need SaaS apps anymore. And the fourth claim is that AI is going to mean infinite competition, that everyone can build an app now. So there’s going to be 10,000 different email service providers and 5,000 different CRMs because everyone can vibe code one and that you’re actually going to have infinity competition so that you can’t possibly compete. And I’m going to take all four of those in order and talk through them and talk through why I don’t believe any of them is the death of SaaS and even why all combined they’re not going to kill SaaS. So let’s start with the custom app and self-hosting your own tools because I feel like this one’s relatively easy to knock down. 

At MicroConf and TinySeed, for example, we pay for, and I’m going to just take a wild guess, 50 subscriptions, 50 SaaS products, Dropbox and Google Workspace and whatever else. Maybe it’s a hundred. It doesn’t really matter, but it’s a lot. Who in their right mind is going to AI code and host 50 SaaS apps? The idea that we would rebuild all of these ourselves is fucking insane. It would be a catastrophic use of our time. And that’s not just because we ourselves are not a software company. If we were a SaaS company, even if we were building software and that’s literally what we do, it’s still a catastrophic use of your time because you should be making far more money specializing in what you’re good at than reinventing 50 tools that already exist and work. I’ve said this on the show before. There are only two legitimate reasons to build your own versions of an app, like to vibe code it. 

One is to save money and that’s only if it’s really expensive. Let’s say something is $10,000 a year, 20 grand a year, and you think you can cut that by 90% or 95%. The second one is customization. No app on the market does what you need it to do. So I want to talk through the saving money. It almost never holds up if you’re growing because you’ve heard me talk about this before and I’m about to record a voice sample that we can just copy paste into future episodes so I can stop saying this like a broken record. But if you are growing as a SaaS company by $1,000 of MRR in a month, we take that multiply by 12 to get the ARR growth, so 12,000. And if you were to sell at a 5X ARR multiple, that’s $60,000 of net worth that you created in a single month from $1,000 of MRR growth. 

And if you sell at a 10X ARR multiple, which is ambitious, I will say that’s high, but I’m trying to give you the idea here. That’s $120,000. So ask yourself, should you spend 20, 40, 60, 80 hours coding something to save you $5,000 a year or $10,000 a year? Or should you put that time and energy into growing your SaaS top line by a thousand MRR to create that 60 or $120,000 of enterprise value? It’s not even close. Building to save money is a rounding error. It’s a bad decision next to just building your actual business. The only time I can think of when this might be worth it is if you’ve been flat for a very long time, your business isn’t actually worth that much and you really are kind of milking it for cash. Some people will do that. That’s the thing. Each of these times I’m going to say, oh, some people will do that. 

And we don’t know how many some is. Is that 2%, 5%, 10% of companies? It’s a small number. This is not 80% of companies. So when we talk about building to save money, there’s going to be a small subset of companies that should do that. In addition, if we talk about building for customization, there will be a small subset of companies that do that, but not that many companies are going to want to build something and maintain it. That’s super customizable. I’m seeing some people do this with CRMs. Maybe that is a particular product category that gets hit really strongly, but is that going to happen to your email service provider? Are you going to really gear up all of the sending infrastructure to replicate an ActiveCampaign or a Drip or a Mailchimp? If it’s sending SMS messages, if it’s doing any type of just the complexity of all of this, it’s like, this is a bad decision. 

(10:29): Don’t do it. And then the last piece of this is in this example, I’m talking about a lot of SaaS companies doing this and we know how to build software, but what about the construction firm and the dentist and the YMCA down the street or the gym, the martial arts dojo, the real estate agent, the psychologist or mental health clinic, the manufacturing company, the beverage factory? These are all examples by the way. Everything I’ve just said are examples of companies that TinySeed portfolio companies serve. Imagine those companies vibe coding and hosting their own app. They get to keep it secure. They get to keep it updated. They get to host it. They get to back it up. They get to own it forever, maintain it, add features. Do you think they want to do that to save, again, what’s the number? Five grand, 10 grand, 20 grand a year? 

They don’t. So that is my rebuttal to the self-hosting fantasy. I think a lot of us on the internet who are technical think a lot of people are going to do things that they really don’t want to do. Claim number two is that anyone can clone your app in a day. And here’s the dirty secret about copying an app. People could always clone your app. This isn’t new. With Drip, we had competitors from, I don’t know, month five after we launched. Every app that I’ve ever built has been copied. In fact, almost every company I’ve built has been copied, including MicroConf and TinySeed. Cloning the software was never the hard part. Getting someone to know you exist, to trust you with their business, to switch off what they’re already using and to stay for years is the hard part. AI writes the code. It does not write your distribution. 

It does not do your marketing no matter what people on the internet think. It does not build your brand. It does not build your reputation. It does not build your customer relationships. A clone with little or no distribution is a folder on someone’s laptop. It always was. The moat was never the code. It was everything you added to the code. I’ve been saying this on the show for 16 years. AI has just made the obvious part the lowest risk part of building. It’s made it faster. So notice what that means. The founders who are panicking are the ones who thought their code was the moat. I hope you don’t think your code is the moat or have ever thought it was a moat. I literally called this out four years ago in The SaaS Playbook and I said false moats in SaaS are features, meaning code. 

They’re false moats. You think they’re a moat, but anyone can replicate your features. AI just allows them to do it a little bit faster. So the founders who were panicking are the ones who thought their code was the moat. They were already in trouble. AI did not create this problem. It just exposes it and maybe makes it a little worse for those folks who aren’t willing to learn how to market and sell. Claim number three is that agents will just do the work. So this one’s more interesting. And the honest answer for me is that agents are going to do some of the work. We come back to this some. How much is that? Is it 2%, 5%, 10%? Agents are not going to do 80% of the work on their own. Some workflows will get absorbed into agents. That’s a real thing. If the product was the workflow, meaning a thin layer moving data from A to B, an agent might eat it. 

So there is a class, a category of apps that is probably going to get absorbed by agents. Now people are going to pay for agents. So if we say SaaS is dead, but agents, which are just software, are going to now be something that you pay for on a recurring basis, that still fits my definition of SaaS. My definition of SaaS is subscription software where software provides the bulk of the value and an agent is still just SaaS. So it’s literally like taking money from one bucket, putting it in bucket B, but those two are still in bucket C. They’re both all contained in a bigger container. That’s how this looks. People will still pay for that. They don’t want to build all their own agents. In addition, agents are only going to eat some of these workflows. I believe most workflows will not get absorbed. 

And the reason is because agents have to act somewhere. The dentist’s agent still needs a system of record, patient data, scheduling, billing, insurance, compliance. They want it to be durable. They want it to be secure. They want it to be accountable and they don’t want to lose data or break the law. And an agent floating in space is useless to them because it needs a backend to act on. So agents aren’t going to kill SaaS. They need SaaS. They actually increase the demand for good APIs, whether it’s CLIs, MCPs, however we want to look at it. Yeah, I get it. There’s some uncertainty these days. But underlying those is still going to be a lot of SaaS that has structured data that’s a reliable system of record. So for a whole category of SaaS, agents are actually going to be a boon, going to be an accelerant, not something that slows them down. 

The future to me is not agents instead of SaaS, it’s agents on top of SaaS. And there are going to be a lot of SaaS founders who are going to build agents into their own SaaS. So it just becomes part of the SaaS subscription. I think it’s going to be a very common model. So if I owned an email service provider, I should probably let other agents interact with it, but should I also build one inside of my own SaaS and potentially charge for it, include it in my subscription, whatever I do? Yeah, probably. And whether they’re using my agent with my product or an external agent, somebody still has to own and run the thing underneath and that’s SaaS and that someone is going to get paid. The fourth claim is that there’s going to be infinite competition. So why should I even start? 

So everyone can code an app now and there’ll be a thousand email service providers or 10,000. Why should I start one? How do you even compete? Two answers of course are it was never about building. It’s always been about distribution, which I’ve already said in point two above. The winners are always the ones who figure out how to market, sell, et cetera. But the second part of this is that I don’t know about you, but the vibe coded apps that I see, the competitors who come in on the, I’ll just say it’s like a get rich quick thing of like, oh, I can vibe code an app, so I’m going to vibe code an ESP because there’s a lot of them. They just seem to drop off the face of the earth very quickly. The people spinning up an app over the weekend with no idea what they’re doing on the tech side, they’re the least experienced entrepreneurs. 

They’re the least experienced people in the market. They don’t know how to get traction. They don’t know how to sell. They don’t know how to support customers. And the vast majority of them I think are going to quit the moment it gets hard because they’re doing it because they think it’s easy. So that’s not the competition I would be scared of. It’s the competition that I think is going to evaporate. More apps getting built is not going to mean more companies getting built because those are two different things. So I’m not here to tell you that nothing is changing. That makes me the mirror image of the doomers. The extreme everything’s dead, everything’s going to change or nothing is going to change. Neither of those extremes make sense. Some things genuinely are going to get squeezed. Consumer and prosumer apps specifically I think are going to take a big hit because consumers and prosumers are so cheap and they’ll blow a weekend coding something, vibe coding it to kill what, $100 a year subscription. 

And that of course is one reason I never focus on B2C. It’s more true now, not less. In addition, I think simple single feature utilities are going to take a hit. They’re just easy to vibe code. The app that someone pays nine bucks a month or 50 bucks a year or even nine bucks one time to convert a PDF to a JPEG or turns a ZIP into a RAR, or I had an SEO keyword tool called HitTail that was kind of, it was like one feature with a few screens. I think you could build that in three or four days now. If your entire product is one feature, I think you’re exposed. And I’m sure there are some specific categories of apps, maybe it is CRMs that take a big hit for whatever reason because they’re overpriced and because you want your own data. That could be a thing that happens. 

But I do not see this happening across the board the way some people are claiming. In fact, here’s the part I think that the pundits or whoever’s mouthing off about this, the doomers, the keyboard warriors get backwards, like precisely backwards. They seem to think AI is going to kill all SaaS or that it’s going to kill the little guy. I think it’s the opposite. The real SaaS apocalypse is the big companies, the big incumbents. They’re the ones who have raised their prices so much, maybe a little too much because they have to grow every quarter for Wall Street. So they have bloat, they have legacy, they have a customer base that’s probably quietly resentful about the last three price hikes. Those are the companies that people are now motivated to replace and AI just lowered the cost of building that replacement. But in my opinion, if you’re a bootstrapper or small SaaS or an upstart, you don’t have that baggage. 

You don’t have the bloat, you don’t have the legacy, you don’t have to grow forever every quarter. You don’t have a code base and a board demanding you to do stuff. You can move fast. You can price fairly and you can go straight at an overpriced incumbent that’s lost the plot. That’s not a threat to you. I think it’s a huge opportunity for early stage bootstrappers. I think it’s honestly one of the biggest opportunities in years. So I think that the haircut isn’t coming for the new startups. I think it’s coming for the big slow moving overpriced incumbents. And startups are actually, if anything, have better tools these days to get to market and compete with the big incumbents. So as I said at the top when I first started this answer, what’s the future of SaaS now that AI allows you to build anything in a weekend? 

The future of SaaS is not fewer companies. It’s a higher bar and SaaS doesn’t need to become something else to survive it. Doesn’t turn into agents, doesn’t dissolve into everybody’s weekend side project. It stays what it is. The future of SaaS I believe is SaaS. And I do think that table stakes features are going to be commoditized faster than they ever have been. You can’t coast on a feature list anymore. You never really could, but it could buy you time. What’s non-copyable these days is what it always was. Owning a market, a brand, a distribution channel, and the trust that you earned by being there for years and keeping data safe, answering support, building a customer base that loves you. If your only asset has been code or features, be nervous. If you’re a commodity, your code just got cheap and so did your competitor’s code. 

But if you are building a company, as I talk about on this show and have for 16 years, this is one of the best times there’s ever been to do it because the stuff you can’t copy, reputation, relationships, your network, your marketing channels, maybe even your audience matters more now I think than it ever has. And the incumbents who forget that are frankly about to find out. So the internet was not dead in 2000 and SaaS is not dead in 2026. The future of SaaS in this day and age is still SaaS. It just belongs to the people who are building real companies all along. So thanks for that question to all who have asked it. And I guess my hope for this episode is I hope you’ll listen to it and absorb it and use it as something to help quell some of the fears that you might be feeling and also to share it with someone the next time that they tell you that SaaS is dead and you ask them for their source and they say, “Oh, it’s some clown who’s never built a company.” It’s like, why would we listen to them? 

Or they say it’s the public markets and we say, “Why would we listen to them? We’re not public SaaS companies.” And frankly, the public markets can be wrong. They are betting on the future, but they can be wrong. And I think that the bootstrappers are in a great position to take advantage of this. If you want to share this episode, obviously you can come to startupsfortherestofus.com and search for episode 842, or you can go to YouTube and search for it. And it does not include full video, but it is an easily shareable link. There are a lot of things that are changing and that scares some people. And when things change, some people think, “Well, that just means everything is dead and we’re just paving over the old and moving on with the new.” And I genuinely do not believe that’s the case. And it’s not just because I’ve been doing SaaS for almost 20 years now. 

Because if there is a new thing, I’d be well poised to jump on it. And especially if that new thing is software that people charge subscriptions for, whatever it’s called, you know that I’m going to be talking about it and you know that I’m going to be paying attention and trying to bring everything that I learn about it to you on this show. So thanks for listening to me today, talking about the future of SaaS in an AI world. I appreciate you listening this week and every week. This is Rob Walling signing off from episode 842.