The short answer
Online communities are a record of people describing problems in their own words, often with the workaround and the frustration attached. The richest sources are niche forums, one-to-three-star reviews of established products, job listings for manual roles, vendor support threads, and posts asking whether a tool exists. Read them to understand problems, not to advertise, and contribute more than you take.
On this page (8)
Why communities beat brainstorming
When people post in a forum or leave a review, they are not being polite to a founder. They are complaining, asking for help or warning others, and they describe the problem in the vocabulary they actually use. That makes community content unusually honest evidence of pain, and it gives you the exact words to use later in your marketing.
It also has limits. People who post are not a representative sample; they skew towards the frustrated, the vocal and the technically confident. A thread with lots of agreement shows that a problem exists for some people, not how many people have it or whether they would pay. Treat community research as a way to generate and sharpen hypotheses, then confirm them in direct conversations.
Five places to look, and what each tells you
| Source | What to search for | What it tells you | Main limitation |
|---|---|---|---|
| Niche forums and subreddits | 'How do you handle', 'what do you use for', 'I hate' | Workflows and workarounds in users' own words | Vocal minority, may not be buyers |
| 1-3 star reviews of incumbents | Repeated complaints across many reviews | Where established products fail specific users | Some complaints are about price or support, not product gaps |
| Job listings | Roles that describe manual, repetitive tasks | Work companies already pay salaries to get done | A role may cover many tasks, not just the one you want to automate |
| Vendor support threads and feature boards | Long-standing requests marked 'planned' or ignored | Gaps the vendor has acknowledged but not closed | The vendor may close the gap at any time |
| 'Is there a tool that' posts | Exact phrases asking whether something exists | Active search, which is stronger than passive pain | The asker may accept a free workaround in the replies |
Reading low-star reviews properly
Reviews of established software, apps and services are one of the most efficient sources available, because the reviewer has already paid for or adopted a solution and is telling you exactly where it fell short. Five-star reviews tell you what to keep; one-to-three-star reviews tell you where the opening is.
Read in bulk rather than one at a time. Collect thirty to fifty low-star reviews for one product and tag each complaint with a short label: 'too complex', 'missing feature X', 'bad for small teams', 'pricing jumped', 'poor support'. Patterns appear quickly. A single reviewer furious about a missing feature is noise; a dozen reviewers from the same kind of business mentioning the same gap is a lead.
Pay special attention to reviews that describe who the reviewer is. 'As a solo bookkeeper, this is far too much software for me' points towards a segment that the incumbent is overserving, which is a classic opening for a simpler, cheaper, focused product.
Job listings and support threads as hidden demand
Job listings are evidence that a company is willing to spend real money on a task every month. Search for roles with phrases like 'manually reconcile', 'data entry', 'compile weekly reports' or 'update spreadsheets'. If many companies of a similar type are hiring for the same repetitive work, a tool that removes part of that work has a budget it can be measured against.
Support forums and public feature-request boards for popular tools are similar. Look for requests with many votes that have stayed open for a long time, and for threads where users share elaborate workarounds such as scripts, templates or chained integrations. A workaround that people bother to share is a sign the pain is real and the vendor is not addressing it.
A simple process for community research
- Pick one niche — A specific trade, role or tool's user base. Broad communities produce broad, untestable ideas.
- Read before searching — Spend an hour reading recent threads to learn the vocabulary and the recurring frustrations.
- Search with problem phrases — 'How do you', 'is there a tool', 'I wish', 'so frustrating', 'what do you use for', combined with the niche's terms.
- Log, do not react — Record each problem in your idea log with the source, the exact wording and the workaround mentioned.
- Count repetitions — Increase a tally when the same problem appears again from a different person or place.
- Move to conversations — For the strongest patterns, ask a few posters politely if they would be willing to talk about how they handle it.
Ethics: do not spam the communities you learn from
Communities exist for their members, not as a lead list for founders. Most have rules against self-promotion and unsolicited direct messages, and moderators remove accounts that ignore them. Beyond the rules, there is a practical reason for restraint: if you burn your reputation in the one community where your customers gather, you have lost your best channel before you have a product.
- Read and follow each community's rules, especially on promotion, surveys and direct messages.
- Contribute genuinely — answer questions you know about, share useful resources, and be a member before you are a founder.
- Be transparent — if you ask about a problem because you are exploring a product, say so plainly.
- Do not quote or screenshot people's posts in public material without permission, and never expose identifying details.
- Do not create fake accounts, post fake questions or seed fake demand. It is dishonest, and it corrupts the evidence you are trying to gather.
- When you do launch, share it where the rules allow and only once, rather than repeatedly dropping links into unrelated threads.
Your action checklist
0/7 doneSaved in this browser only. Nothing is sent anywhere.
Questions
Is Reddit good for finding startup ideas?
It can be, especially in niche subreddits where practitioners discuss their work. The value is in reading how people describe problems and workarounds. The limitation is that posters are a self-selected group, so a popular complaint does not prove a large or paying market. Use it to form hypotheses, then confirm them by talking directly to people who have the problem.
How do I use app reviews to find business ideas?
Collect a large batch of one-to-three-star reviews for an established product, tag each complaint with a short label, and look for repeated patterns from similar types of users. Repeated gaps for a specific segment suggest an opening for a focused alternative. Separate complaints about price or support from complaints about missing capability, since they point to different opportunities.
Is it okay to message people from forums about my idea?
Only if the community's rules allow it, and only in a way you would be comfortable receiving yourself. Be clear about who you are and why you are asking, keep it short, and accept no as an answer. Many communities forbid unsolicited messages entirely, in which case ask publicly whether anyone would be willing to chat, if that is permitted.
What does an 'is there a tool that' post really tell me?
It shows someone actively searching for a solution, which is stronger evidence than general complaining. Read the replies too: if others suggest a free workaround that satisfies the asker, the opportunity may be thin. If replies are full of people saying they want the same thing and nothing fits, that is a much more promising signal worth logging and investigating.
Finished? Mark it so your paths and the library remember.