Look for the workaround
Watch for spreadsheets maintained by hand, information copied between tools, repeated requests in a team chat, and tasks postponed because setup takes too long. These are leads for investigation, not proof that someone needs a new app.
Write a problem statement using this template: When [specific person] needs to [task], they currently [workaround], which causes [observable consequence]. For example: when a small contractor prepares a quote after a site visit, they reconstruct notes and measurements from messages, which delays sending the quote.
Ask about the last time it happened
A conversation about a recent event is more useful than asking whether someone likes your idea. Let the person walk through the process and show an example if they are comfortable doing so. Avoid collecting customer records or other private information you do not need.
- When did you last do this, and what triggered it?
- What steps did you follow, and which step was difficult?
- How often does this happen?
- What have you already tried? Why did you keep or abandon it?
- Who would decide to change the process?
Separate a complaint from a buying problem
Some frustrations are real but too infrequent to justify a separate subscription. Others are better solved by a setting in an existing tool, a template, or a service. Investigate these alternatives before deciding the answer is software.
Record the cost in the person's own terms: time spent, mistakes corrected, work delayed, or an existing budget. Do not turn an estimate into a precise market-size claim. A person already investing effort in a workaround is worth speaking with, but that still does not establish willingness to pay.
Choose one audience and one job
Narrow the problem until you can find the same situation in several independent conversations. ‘Help small businesses’ is difficult to test. ‘Prepare a first quote from site-visit notes for independent contractors’ gives you a specific workflow and a place to start.
Browse the problems on WorthBuilding for starting points, read their current workarounds, and investigate the people who encounter them. The initial collection includes team-contributed ideas and AI vote points, disclosed in the site policy. Treat those as exploration prompts and gather your own customer evidence.
Write down the next experiment
Finish with a small test: manually prepare the output, show a prototype to someone with a recent example, or observe their current workflow. Define what you expect to learn and what result would cause you to stop. If you cannot explain how to reach the people affected, investigate distribution before investing in a full product.
Explore problems to investigate
- Small business software ideas
- Developer tool ideas & engineering problems
- Productivity app ideas & everyday friction
