TL;DR: The best SaaS ideas come from places where people already complain, not from brainstorming sessions. Below are ten real sources, ranked roughly by signal quality, plus the three filters that separate a real idea from noise.
The filter to apply everywhere
Before the list, the test that matters at every source:
- Intensity — does the problem actually hurt, or is it a mild annoyance?
- Frequency — does it recur, or was it a one-off?
- Willingness to pay — are people already paying for workarounds?
An idea that passes all three across more than one source is worth your attention. One that only shows up once, mildly, is not.
10 places to find SaaS ideas
- App Store & Google Play reviews. One- and two-star reviews are blunt feature requests in disguise: "I wish it could..." and "why doesn't it..." Reviews across many countries reveal gaps the original maker ignored.
- GitHub Issues & Discussions. Developers describe problems precisely and often propose the fix. Long, active issue threads are concentrated demand.
- Stack Exchange / Stack Overflow. Repeated questions with thousands of views are unmet needs people actively search for.
- Hacker News. "Ask HN" threads and comment sections surface founder- and developer-grade problems, plus brutally honest reactions.
- Niche forums & Discourse communities. Vertical communities (for a specific industry or tool) are where the unglamorous, high-budget problems live.
- Reddit and federated alternatives. Communities vent in detail; "is there a tool that..." posts are direct demand.
- Product Hunt comments. Launch comments reveal what existing products miss and what adjacent problems remain.
- Support docs & changelogs of incumbents. What big tools refuse to build is a map of underserved niches.
- Job boards & remote-work sites. Repeated roles and tasks signal manual work that software could replace.
- Your own workflow. The friction you hit daily is a source most people overlook — if it annoys you and others, it's a candidate.
Turning sources into a shortlist
Browsing ten sources by hand is thorough but slow. This is exactly what a pain-point database automates: PainHunt collects complaints from 24 of these kinds of sources and scores each one for commercial potential, so the three filters above are applied for you and you start from a ranked shortlist. See where the data comes from for the full source list.
From idea to evidence
Finding an idea is the start, not the finish. Once you have a candidate, validate it: confirm the problem is real and that people will pay. Read how to validate a startup idea, or run a candidate through the Idea Validator.