Competing with giants, the end of SaaS & other listener questions

0:00Welcome to Rework, a podcast by 37 Signals about the better way to work and run your business. I'm your host, Kimberly Rhodes, joined by the co-founders of 37 Signals, Jason Fried and David Heinemeier Hansson. This week, we're going to knock out a couple of listener questions. I'm going to start with this one, which was an email from Dan. Jason, you emailed Dan directly with your response, but I thought we would tell our Rework listeners what you have to say. So Dan said, my name is Dan. I'm an avid listener to Rework as well as a big fan of Basecamp, Pei and Omachi. I'm at the early stages of building my own business and would love your thoughts on building a product in a competitive market.

0:35In my space, there are competitors, but I'm confident I can build something better. I wonder how much better does it need to be? As 37 Signals have built products competing with giants like Jira, Gmail, and Apple, I'm curious if that scares you. Do you worry about the big players? That's from Dan.

0:54David Heinemeier Hansson:Sure. I mean, look, you're not going to beat Apple or Gmail at their game because they have massive market share. They're massive companies and they're in a different category. I think if you're Dan, like you're not really competing against them. What you're trying to find are your own customers and carve out a business that's sustainable for you. So you're competing against your costs. You're competing against the ability to find some customers and enough customers to build a great product for them and for yourself. Like you got to start there. You can't be like, I'm going to take down Google.

1:26David Heinemeier Hansson:I mean, I guess you could, but it's probably a bad idea. I mean, in general, my sense is like for Dan, what I basically said is like, focus on your own thing. Like this idea that I'm going to, what do I need to do to take them down? You got to build something first, whatever happens happens after that, but don't think about what they're doing. Think about what you want to do, why you need to exist, why customers can't find the thing that you want to make and how to differentiate yourself and stand apart and build something that's tight, good, and just right for them and for you. That's all you can do to start.

1:57David Heinemeier Hansson:Whatever happens down the road happens. Day one, day two, year one, year two. If you do take down Google, it's not going to happen on day one or day two or year one or year two anyway. Maybe in 10 years, you've got something for that. Right now, just nail it. Get something good out there. Put something out there in the world. Get a customer. Get a second customer. Get a third customer who's willing to pay for what you've made. That's the real challenge you're up against. It's not how do you take down the big guys. I think you got to start also with developing something, making something that you would pick over these other alternatives.

2:30Because as Jason says, they have a lot of advantages. So what you're coming up with has to be distinct enough that at least you would pick and there's a good reason why you would pick it because it's a different spin on it. Now, Now, it doesn't have to be 180 degrees different, but it has to be different enough that someone is going to see a reason to pick you. Because on a label playing field, if it's just Red Ocean, you're selling something that's identical with what they're coming out with. Like, yeah, you're going to have an uphill battle, right? So you got to have some differentiation. And that differentiation has to be the reason why you at least would pick this if someone else had made it.

3:12Not yourself. Now, it's pretty hard to put yourself out of your own bubble in that regard, but I find it's actually a really good exercise to do. You're just like, would I use my own credit card for this? Would I pick this option over whatever the big established players have? And then once you have something that is distinctive enough, you have to carve out the distribution. Like, how are people going to find you? What are you going to say that sounds different, that is different than those setups? The worst thing you can do is falling into the trap of trying to puff yourself up and like, I'm also kind of Google.

3:50I'm also kind of a big software company that's going to do everything for you in all the ways. Now, you have to veer in the other direction and double down on everything that they can't do. As we've talked about so often, even at 60 people at 37 Sickles, we can do a lot of things that a Google could never do. Like Jason, replying directly to you on a question like this, that's not something that's accessible to CEO of a 60 ,000 person company, right? So you have to lean into all of those aspects of being small that people are going to resonate with. Because there are plenty of folks out there who actually, if given a reasonable, proper, distinctive alternative, would prefer not to buy from Google or Apple or any of the others.

4:34So being there for them, even if it's a niche, is a great place to be. Okay. I'm going to go to this question, AI-related question from Arun. He says, hello, Jason and David. I've been a regular listener of the show for the last couple of years, and I have a question with regards to AI and the buzz around the whole shift from SaaS to AI agents. Is this really the end of SaaS business as we know it? Okay. Can we first talk about that? Like, are people thinking that agents are replacing SaaS? I think there are certainly people thinking that they will replace SaaS. And there are businesses that are replacing SaaS with homegrown solutions.

5:11But I don't think it is in any way, shape or form a majority thing right now or even barely early adopter. This is pioneer level stuff on the adoption curve for anyone to take something that they're already using and going like, no, I want to save some money here, so I'm going to build my own. Triply so at the low end. Almost all of the SaaS replacement I've seen has been from folks who have an intrinsic interest in doing agent exploration work. And therefore, they go like, oh, here's a piece of software I use. I should try to build my own version of that. But you're not doing it really because you're going to save on your bill if you're a small company.

5:53You're doing it just because you want to play with agents. And I think that's totally fine. That's totally great. And then what's happening is that there is some early inklings at the top end, the kind of companies who might pay Salesforce$600 ,000 a year for a system that they hate. That is absolutely a danger zone, I think, right now in enterprise SaaS. And I've seen that play out in a number of large companies, Shopify being one of the examples I've talked about before, where they literally took that function, the CRM part of how their sales force was interacting with their leads, and they just built their own.

6:33Was it because they were going to save some money? I don't think that was the motivating reason. The motivating reason was that what they were using was crap. And I do think if you have a crap piece of software that's just there because it's entrenched, because it feels too hard to take it out, yeah, you're going to be first in line for the danger zone here. If those two things can conspire, like why are we paying someone, let's say again, $600 ,000 a year for a piece of software everyone loathes to use? Could we do something better? So at our end of the market, I'm not seeing a lot of evidence.

7:10And I think that end of the market goes quite high. It's not only the small team, it's also the mid-sized team. And I know Jason has a whole spiel on the fact that most people actually don't want to build their own stuff. And I think with SaaS, it's not just that they don't want to build it, it's also that they don't want to run it. And we got an abject lesson in that with once. We sold SaaS software for a single fee where someone could save a lot of money on, let's say, a Slack installation by replacing it with Campfire. And the uptick was not enormous. And I think one of the reasons why that was was that they don't want the obligation.

7:46They don't want the responsibility of having to care for and tend to a service and manage the data and worry about backups and all that other stuff. And has that changed? I don't think so. Will it change? I think you're foolish if you don't think that the AI wave is going to drastically rearrange the software landscape. And I don't think there's any part of that landscape that is safe, that's never going to be touched, that's never going to have challenges from agents building things. I think that's silly to think that. But what's the curve? Are we talking six months from now? Are we talking two years from now?

8:27Are we talking five years from now? I think we're closer to the long end of that tale when it comes to what most people would traditionally think about SaaS, not these huge honking enterprise installations, which, by the way, are already slob code, right? The thing with a Salesforce installation is it's not off-the-shelf software. That's part of the pitch is we're going to take this product and then we're going to send in some consultants or Ford engineers to really tailor to your specific setup. And that's one of the reasons why these products often just aren't that nice to use because they're kind of built in a bit of a sloppy way.

9:06So if you're going to build a slop like that, yeah, just go straight to the source and get it from an agent.

9:10David Heinemeier Hansson:I have like a bunch of unrelated things to say, and David hit on most of them, but a couple of things. First of all, there's already risk that people are going to leave your product for another product that exists. So the idea that there's been no competition and there's no alternatives, that's what I hear a lot when people are like, well, is AI going to take my customers? Well, your competitors might take your customers. Your product may be bad and they may leave anyway. There's a million reasons people might leave or might come to see you. So let's start there. It's not like there is no risk currently.

9:39David Heinemeier Hansson:Yeah. We have customers who leave every day. We have customers who come back. We have customers who leave other products and come to us. There's a constant churn of people moving around anyway, regardless of AI. AI is another option. If someone wants to build their own tools, that's another way for someone to leave your product. But the other reasons are already there. So let's start there. Second, like David mentioned, I wrote this piece on, I just don't think people are going to want to build bespoke software. Most businesses are not going to want to build bespoke software. Building software takes a lot of work, even if it's become way easier.

10:09David Heinemeier Hansson:someone still has to maintain it, someone has to think about it, someone has to think about the workflows, someone has to adjust things. If that person leaves, then what happens? And people are busy already with their own things. It's understandable that like online, on X, on places like that, where everyone's talking about software, people think everyone in the world is talking about software. Most people just don't care about software. They care about the plumbing business they run, the electric business they run, the whatever business they run. And yeah, could they build their own thing, they probably could, but why would they?

10:39David Heinemeier Hansson:Now, this also gets back to David's point about price. So like David said, most of the stories I've been hearing are when people are trying to potentially save a million bucks a year or whatever on something they didn't like to begin with. So when I look at our business, a lot of people pay us 50 bucks a month. That's 600 bucks a year. If you look at how much it would take to build Basecamp, even if you're using AI and you took one or two or three employees to kind of do it and mess around with it, whatever, it's going to cost you thousands and thousands and thousands of dollars anyway, and opportunity cost and things they could have been doing otherwise to save 600 bucks a year.

11:17David Heinemeier Hansson:And then like you don't have something that's quite right. And then you've got to keep maintaining it and things might break. And then who do you blame? A lot of buying products is about actually ultimately be able to blame someone and put someone else on the hook to take care of you. That's why you pay them money. So, you know, there's all these reasons why it doesn't make sense, especially in the low end. But again, as David said, like, we'd be foolish to think that everything is safe and nothing's going to change. Of course, things are going to change, but also we're going to change and we are changing.

11:46David Heinemeier Hansson:We're able to build things faster and better. And we have 25 plus years of experience understanding what software is about. It's not just about compiling code and having software do things. It's about what do people actually want to do? Why do they want to do it? How do they want to do it? How do these things all work together? How do these pieces fit together? You can dream up all sorts of things, but that's not what software has ever been about. Great software, I should say. Great software is still about a cohesive concept, an idea, a product, borders, boundaries. That's what this is all about.

12:19David Heinemeier Hansson:So no, are we all immune? No. Are things going to change? Yes, absolutely. But you got to think about where you are. This is not a monolithic thing. Software is not a monolithic thing. There's different levels or different prices and the kinds of companies and customers you have may or may not be interested in getting involved in doing their own stuff at all. We all hire contractors. We all hire vendors. We all hire people to do things for us that we could do ourselves, but we don't want to because it makes more sense to hire them to do those things. I have a housekeeper that comes once a week to clean our house.

12:51David Heinemeier Hansson:I could clean my house that day too, but like, what does it cost me to take a day off of work from doing that? Or my wife and I doing that? Or like we keep the house kind of clean, but we We also have someone who comes by and cleans the house. Like, it just makes sense to do that, right? So if you just pile up all the things in your world that you do or you kind of have other people do or whatever, it just makes sense sometimes to buy something off the shelf or from a company that you like that makes a great product that supports what you do. So you don't have to do it yourself. These things remain true, will always remain true.

13:22David Heinemeier Hansson:But yes, of course, things are changing. I think a great illustration of this is the episode of The Simpsons, where Homer is given the opportunity by GM to design his own car. And individually, all his little ideas for, oh, there should be extra cup holders here, there, and everywhere, they're probably fine. And that's what a lot of customer feedback we get, too, is individually, everything makes sense in that moment, narrowly for that person at that time. But you take it all together and you end up with a homeless car, an absolute abomination of lack of vision, lack of cohesion, lack of something people want to use.

14:01Most people have just not thought that carefully about the problems that they have and may not be blessed with the competencies or even the talent to come up with solutions to those problems. They know their problems. I think this is what you'd always have to respect as a product builder is that the customer knows their problems. They do not know the best solutions. That's why they're coming to you. That's why they're hiring you. So if they are in charge of coming up with solutions to their own problems, they're going to develop a better horse or a faster horse, not the Model T. Now, I think the tipping point is actually not AI doing what a customer tells the AI it should do and makes the software that they want to make.

14:47I think the real tipping point is once the AI tells you the kind of software that you should have. It looks at your problems and places itself in the role that we have as problem solvers and product developers. We're not there yet, or I haven't seen that at least be very successful in any shape. all the best use cases right now that exist in the world for AI aiding development is someone with taste, vision, problem space, knowledge, feeding all that in and steering a bunch of agents to build that for them. If you flip it around and say like, well, I don't know what I want. I've not seen good things come out of that.

15:28And I've tried quite a few times where I'm like, let's just see where it goes with this if I just let it drive. And as of yet, it's not. But That's the part where give it another five minutes, six months, two years, five years. Who the hell knows when this whole thing is going to tip in that direction where it's AI telling us what the world should look like. I also think it's interesting that I've seen so many unattractive products, unattractive apps. And yes, it could be programmed by AI, but there is a design element of taste that I don't think we've gotten to quite yet. I think that's actually the main charge against slob, which is this framing for AI just producing a bunch of stuff.

16:12No, the slob is not coming from the agents. It's coming from you. It's coming from you, the instructor of the agents, that you don't have the vision and the eye for developing something good. And why would it be otherwise, right? Like this is why you have product designers sit down and think carefully about building things. It's not just about making something happen. That is the big change that has happened. And I do think that that's perhaps also the other flank of this that's very interesting. There are actually quite a few people in this world who are quite good at having some taste and some vision and an eye for where things should go, but did not have the talent or competence to do implementation, did not have the funds to do implementation.

16:55So at least for the foreseeable future, I think that's the bigger threat to anyone running a software company is that someone else, another human out there is suddenly capable of making their ideas come true. I'm actually quite certain that if AI had been around in, say, 2000, when Jason put the post out for a programming question that I answered and we started working together, I'm not so sure I would have ended up moving to Chicago. Because Jason would have been able to build a lot of things by himself. Because at that time, that was the constraint on his ability to create software. He had all these ideas and the designs and how it should go, but was lacking the mechanism of turning those into software.

17:35Now a lot of people have suddenly gotten a really good mechanism for that. Okay, let's go to this question. It's an email from Alex. Hi, Jason and David. It seems that the two of you promote your business through your own personal brands, most notably on X, although I'm sure you're also probably on LinkedIn. I'm starting a business soon, and I always thought that I wanted to keep my personal social accounts like X, Instagram, and TikTok completely separate from my business. But seeing that you two are using your personal brands to promote your company makes me wonder if this is the right play.

18:06Would love to hear your thoughts from Alex.

18:10David Heinemeier Hansson:I mean, I don't think there's such a thing as a personal brand, by the way. I've always been bugged by that whole thing. It's just me talking on the internet. Where do I work? I work here. What do I talk about? I mean, but that is your brand. Like your style of writing, your personality, the things you're saying. Do you know what? Let me just interject. Brand to me has the same slot like content. It's one of those big wrapper boxes that just people talk about in the abstract, and it's really fucking annoying. I'm not. Thank you. Right? Yes. It's just as in a general sense, this idea that we should be talking about people like brands or products, even as products.

18:51I hate that too. As soon as you get more and more generic and upstream from that, you lose all the essence of what makes something interesting. And certainly if you reduce a person to a brand, now you've really cooked out all the nutrients. Sorry.

19:06David Heinemeier Hansson:No, I agree. Thank you for that. I mean, I just, just like, I talk about what we're doing primarily. I don't share that much personal stuff on online. Like David and I run the company. Like this is our company is the kind of stuff we talk about. David's talking about Amachi a lot lately. I'm talking about, like, we just talk about the stuff we're interested in. And that just, for the most part is business related, not completely. We talk about cars and watches and some political stuff and some other points of view, whatever. But we work here. That's what we do. That's what we talk about. And I would encourage you to do the same.

19:36David Heinemeier Hansson:I wouldn't think about like, is this a personal brand post? Just what's on your mind? Share it however you want to share it. People will like or they won't like it. And just don't overthink it. That's just it. Just don't overthink it. But have something to say. And people will like it or they won't like it. It doesn't really matter at the end of the day. Like you just have to be yourself and whatever you're attached to will get some exposure. And if it's a business, we'll get exposure. If it's your personal stuff, your personal stuff will get exposure. And that's that. I mean, I don't know what else to do there, frankly.

20:05David Heinemeier Hansson:That's my own little silly take on it. I will say the other problem I have with this notion of brand is it's very manufactured. Brand is something you make, is something you craft, is something you sit down and design. And that is the opposite ends of what I think actually is resonating and working is authenticity. Now, authenticity in itself has become a fucking brand word that barely means anything, right? But this notion of being yourself in a way that someone can tell isn't just all about the objective, right? As soon as you talk about personal brand in relation to promotion, you talk about playing this role.

20:49And I think, if anything, folks online have gotten quite good at sniffing that out. Who's inauthentic? This is why you mentioned LinkedIn the way you did. Because people fucking hate LinkedIn. Because LinkedIn is nothing but people trying to project this manufactured version of themselves onto other people who don't give a fuck about what that is, right? LinkedIn is really the seventh ring of hell when it comes to this projection of fake personas. Now, the other part of that is now literally fake because half of the posts are just AI slob being filtered through some prompt about how I need to position myself in the marketplace of ideas.

21:33I mean, Jesus, shut up. Right? And I think it's actually one of the reasons I post on LinkedIn. And I just post the same stuff that I post anywhere else on LinkedIn. But I remember Joe Rogan had this wonderful imagery of how he used to use Twitter. And he called it, you just got to get in there, man. Throw the grenade and then run. As like, don't interact with it at all. And I kind of feel bad about that with LinkedIn, but also not really. Because the blast radius of that grenade in there is fucking 70 % just filler nonsense. Now, again, I've actually in the past been surprised by the audience and individuals we've been able to reach on LinkedIn.

22:16It is just a different crowd, and I can also appreciate that. But you've got to get away from this manufactured personality. It's just nasty to be trapped in this because it's just not who you are. And if who you are can't sell, then try to become a better person, not a better projection.