Why idea lists do not work
A published list of startup ideas has been read by everyone else who reads such lists, which removes the only advantage an early entrant has. Worse, the ideas on it are usually generated from what sounds plausible rather than from what someone actually asked for, so they carry no evidence with them — you cannot check them, and you cannot find the person who needs one.
The useful version of the same document is a list of problems with the evidence attached, which is a different artefact entirely: it does not tell you what to build, it tells you who is stuck and where.
Where good startup ideas actually come from
Three sources, in descending order of reliability. First, a problem you have personally and have already built a workaround for — you are your own evidence, and you know the market's language. Second, a problem you have watched someone else fight repeatedly, close enough to see the cost. Third, a problem strangers describe in public often enough that the pattern is unmistakable.
The third is the only one that scales, and it is the only one that can be read systematically rather than waited for — which is what this platform exists to do.
Turning a complaint into an idea you can test
Take one complaint and answer four questions in writing. What exactly did this person try? What did it cost them when it failed? What would they have accepted instead? And what are they doing today, now that they have given up? The fourth answer is your real competitor, and it is usually a spreadsheet, a manual process or nothing at all.
An idea that survives those four answers has a shape, a price ceiling and a named first user. An idea that does not was never an idea, only a topic.
What to ignore
Ignore anything whose appeal is that the technology is new. New technology changes what is possible to build, not what people need, and a demand that only exists because a tool exists disappears when the next tool does. Ignore ideas whose entire premise is that an incumbent is bad; being better is not a business unless the incumbent's users are actively leaving. And ignore markets where every complaint ends with someone recommending a free alternative that works.
From idea to first customer
The person whose complaint you read is reachable. They wrote in public, under a handle, on a dated post. Replying to them with a specific question is the cheapest customer research in existence and the step most people skip in favour of building for six months. If nobody replies, that is information too, and it cost you an afternoon rather than a year.