Micro-SaaS Ideas Are Everywhere. Here's How We Pick Which to Build
Our filter for choosing micro-SaaS ideas worth building: markets we can reach, problems that recur, and products one person can run.
By FGA Labs
The hardest part of running a studio is not finding ideas. Ideas are a commodity — every support ticket, every clunky workflow, every 'why is there no tool for this' moment generates one. The hard part is saying no to the ninety-five percent that would consume a year and return nothing. Over time we have compressed that judgment into a filter with three questions.
Can we reach the customer without a miracle?
This is the first question, before product, before pricing. Some customers congregate — in search queries, in communities, in app marketplaces — and some are scattered across the economy with no watering hole. A mediocre product with a clear distribution channel beats a brilliant product that requires cold outbound to strangers. If we cannot name the specific place the first hundred customers will come from, the idea goes back on the shelf no matter how elegant it is.
Does the problem recur?
Subscription software needs subscription problems. A tool that solves something once — however painfully — earns one payment and a churn event. We look for problems with a heartbeat: weekly reports, monthly reconciliation, every-deploy checks. Recurrence is what turns a utility into a business, because the product re-earns its subscription every time the problem comes back.
Can one person run it at scale?
- Support load: does each new customer bring questions, or does the product explain itself?
- Operational surface: no human-in-the-loop steps, no content treadmill, no marketplace to police.
- Failure tolerance: when it breaks at 2am, can it wait until morning?
This last filter is the one people skip, and it is the one that determines whether a portfolio compounds or collapses. Every product we run has to coexist with all the others. A single high-maintenance product does not just underperform — it taxes everything else we operate.
We are not looking for the best idea. We are looking for the best idea we can distribute, that recurs, and that runs itself.