What a solo SaaS actually costs to run per month
Indie builders love to share revenue screenshots. Nobody screenshots the bill. So here is the boring side: what it actually cost, per month, to keep one of my solo projects running, before it had a single paying user.
It was a small web product. Nothing exotic. Roughly where the money went, monthly:
- Hosting and the app: about $20. The Pro tier of my host, needed once you cross the free function limits and the bigger builds.
- Database, auth, storage: about $25. The Pro tier again, because the free one tops out fast once you have real tables and real auth.
- Rate-limit and cache layer: free at my volume. One of the few genuinely free pieces.
- A small always-on worker: about $5. A separate little service doing a job the main app could not.
- A usage-billed data provider: a few dollars and climbing. The kind of cost that starts tiny and scales with success, which is its own trap.
Call it $53 to $55 a month, all in.
Fifty-five dollars is nothing, until it is everything
Stated once, that number is trivial. A nice dinner. The problem is that it is not stated once. It arrives every month, automatically, whether or not anyone is using the thing.
For a product with revenue this is the cost of doing business and you barely notice it. For an MVP that has not launched, or has launched to crickets, it is a slow leak. Fifty-five a month is six hundred and change a year, out of your own pocket, for the privilege of waiting to learn whether anyone wants what you built. I have paid that tax on more than one project that did not make it. The infrastructure does not care that you have no users. The invoice arrives anyway.
The line that scales with success is the dangerous one
Most of the bill is flat. The piece that scared me was the usage-billed provider, the one that costs almost nothing today and grows with every user. Flat costs you can sit on. A cost that scales with usage but not yet with revenue is a different animal: the more the product works, the more it bleeds, right up until you have figured out how to charge. If your unit economics are upside down, growth is not the goal, it is the threat.
What I do about it now
- Stay on free tiers until something real forces you off. Every Pro upgrade should be triggered by a limit you actually hit, not by "might as well."
- Know which costs are flat and which scale. The flat ones are a subscription to your own hope. The scaling ones have to be covered by revenue before you let them grow.
- Put a number on the runway. Fifty-five a month is some amount before you have to either charge or shut it down. Decide that amount up front, while you are calm, not later, while you are attached.
- Count the dead ones. The real cost of a habit of building is not one bill. It is all the small recurring bills from all the things that did not work, quietly overlapping. Turn each off the moment a project is done. I am not always good at this.
None of this is a reason not to build. It is a reason to know the number. The revenue screenshots are fun. The bill is the part that decides how many shots you actually get.