How do I find a real problem worth solving instead of a solution looking for a problem?
Hunt problems, not ideas. Real problems show up as complaints people repeat, workflows they hate, and expensive workarounds they already pay for. Mine those directly: read reviews and forums for recurring gripes, watch how a specific job actually gets done, and look for duct-taped spreadsheets and manual steps. A problem worth solving has a named sufferer who is already spending time or money to cope with it. Attach the who-pays before you design any solution.
A solution looking for a problem starts with "wouldn't it be cool if"; a real business starts with "this is broken and I am already paying to deal with it." The difference is a sufferer who exists before your product does. So do not invent, observe. The best problems are boring, repetitive, and annoying enough that people have built workarounds, because a workaround is proof of both the pain and the willingness to spend on relief.
Use problem-first hunting from the idea-generation-techniques guide. Read one-star and three-star reviews of tools in a space (the three-stars are gold, they describe exactly what is missing). Sit with someone doing a real job and note every manual, repeated, or dreaded step. Search forums and communities for phrases like "I hate that," "is there a tool for," and "how do you all deal with." Each recurring complaint is a candidate.
Many of the strongest entries in our Idea Bank were built exactly this way: from a documented pain with the payer already attached. Review-response management exists because owners hate replying to reviews; tenant screening admin exists because landlords dread the paperwork; import-compliance concierge exists because first-time importers fear getting it wrong. In each, the problem came first and the who-pays was obvious.
Qualify before you commit. For any pain you find, confirm a specific buyer, that they currently spend time or money coping, and that they would pay you to remove it, then score it on the Idea Validation Scorecard. If you cannot name who hurts and what relief is worth, it is a solution in search of a problem.
Prompts to try
Copy these into ChatGPT or Claude to go deeper.
In [industry/role], list 10 recurring complaints, dreaded tasks, or expensive workarounds people already tolerate, with who suffers from each.
I want to find problems worth solving for [audience]. Give me a research plan: where to read complaints, what phrases to search, and how to log patterns.
For this pain [describe], confirm it is worth solving: who exactly suffers, what are they spending now to cope, and would they pay to remove it?