Start with the problem, not the product
Write down who experiences the problem, when it happens, what it costs them and what they do today. If you cannot describe the current workaround, you may not understand the problem well enough yet.
Interview around past behaviour
Ask what happened the last time the problem occurred, what they tried, what they paid for and why the workaround was unsatisfactory. Avoid leading questions such as “Would you use an AI tool that…?” because people are generous with hypothetical approval.
Look for existing spend and effort
A spreadsheet maintained every Friday, a virtual assistant hired for repetitive work or several tools stitched together can all signal real demand. The stronger the current cost or inconvenience, the clearer the value proposition can become.
Test the offer before the full product
A landing page, clickable prototype, manual concierge service or narrow AI-built MVP can test the promise without months of development. State the outcome clearly and ask for a meaningful next step.
Set a decision rule
Before testing, decide what would make you continue, change direction or stop. This prevents enthusiasm from redefining success after weak results.
When the evidence is strong enough to build, move to building the MVP rather than expanding the concept indefinitely.