Why I decided not to build another SaaS product
When I started building my own software products, SaaS seemed like the obvious business model.
Recurring revenue is attractive, and users don't have to think about installation, hosting or updates. But the more I considered what running SaaS would actually mean for me as a solo creator, the less obvious the decision became.
I work independently while also having a full-time job. A SaaS product wouldn't just mean building the application. I would also be responsible for hosting, infrastructure monitoring, security, updates and ongoing customer support.
I realised I was more interested in building software than operating an online service around the clock.
That pushed me towards a different model: self-hosted source-code products. With applications such as My Budget and Captain's Toolkit, the buyer can host and modify the software rather than depending on me to operate it indefinitely.
There is a trade-off on the buyer's side too.
With SaaS, much of the responsibility belongs to the vendor. With self-hosted software, ownership also means dealing with hosting, backups, security and maintenance. More control doesn't remove the work. It moves the work.
I don't think that makes one model better. They create very different relationships between the person building the software and the person using it.
For me, the self-hosted model fits the kind of products I want to create and the way I can realistically work.
I wrote more about the wider Custom Software vs SaaS question here, if useful: https://cordinant.com/blog/self-hosted-vs-saas-what-is-the-difference
For other solo founders: how much did the ongoing operational responsibility of SaaS influence your decision to build or not build a subscription product?
I think this is an important reminder that being a founder doesn't mean you have to keep building SaaS products forever.
Sometimes the better decision is to step back and ask whether the market, problem, distribution, and business model actually justify another product. Building has become easier with AI, but that makes product selection and customer validation even more important, not less.
There is also a lot to learn from deciding not to build. Understanding why an idea isn't worth pursuing can save months of development and create better judgment for the next opportunity.
For me, sustainable SaaS growth starts with solving a real customer problem, not simply having the ability to build the product.
What made you realize that not building was the better decision?
The ops argument is real, but the trade might not be ops for no ops. It could be funded support for unfunded support.
With a subscription, the person emailing you in month fourteen because a dependency broke is paying you that month. With a one time source code sale they are not, and self hosted buyers tend to generate more of those emails rather than fewer, because the environment is theirs and you cannot see it. You have removed hosting and monitoring, which are automatable and get cheaper per customer, and kept the support tail, which does neither.
The shape that seems to survive this is selling the code perpetually and the updates as a renewable year, the way JetBrains does. You never touch a server, the buyer keeps working software forever if they stop paying, and the support tail has revenue attached to it.
Genuine question though: how are you planning to find buyer number one hundred? That is the part the recurring model was quietly paying for.
That's a fair point about support. I'm still early enough that I don't know how big the long-term support cost will be. Since I work alone and also have a full-time job, keeping the operational side manageable is actually one of the reasons the self-hosted model appeals to me. I can't realistically be responsible for running and monitoring customer applications every day.
I do like the idea of selling the source code perpetually but charging separately for future updates or support. That's something I may consider as the customer base grows.
As for buyer number 100, I'm still working on buyer number one. :)
Right now I'm trying SEO and blog content on my own site, Reddit and Indie Hackers, and improving my product pages and documentation. With limited time, I'm trying to find a few channels I can manage consistently rather than be everywhere.
Building the products is still the more predictable part. Finding a repeatable way to reach buyers is what I'm learning now.
Who do you expect the typical buyer to be—a developer comfortable self-hosting, or someone who mainly wants the finished app? That distinction seems like it would shape both the product and how much installation support lands back on you.
I'm mainly thinking about developers, freelancers and small agencies who are comfortable with hosting and want source code they can use or modify.
I'm also a full-time employee, so the time I can spend on my own products is limited. That's one of the main reasons I chose this model. With SaaS, I would need more time not only for development, but also for hosting, monitoring, updates and ongoing support.
I still provide documentation and an installer to make setup easier, but I expect buyers to have some basic technical knowledge. If I targeted people who just wanted a finished app with everything managed for them, I think I would quickly end up with many of the same responsibilities as running a SaaS.