Find a problem worth solving

Start from a painful, frequent problem that a specific group of people already spends time or money on. Ideas are cheap; evidence of pain is not.

Also called hair on fire problem, pain point

What: A problem that hurts enough, and happens often enough, that people already try to fix it with workarounds, spreadsheets, hired help or other products.
Why: People pay to remove pain, not to get a nice idea. Frequency matters because a problem that happens once a year rarely supports a subscription.
Example: Freelancers chasing late invoices every month is painful and frequent. Planning a once-in-a-lifetime event is painful but rare.

Watch out

Starting from a technology you want to use and then searching for a problem it could solve.

Also called Niche, ideal customer profile, ICP

What: The specific group you build for, described precisely enough to find them: role, situation and where they spend time online, such as "solo physiotherapists who book patients by WhatsApp".
Why: A narrow customer makes every later step easier: you know whom to interview, what words to use on the landing page and which community to launch in.

Watch out

Choosing "everyone" or "small businesses", which gives you nobody to talk to and nothing specific to say.

Also called founder-market fit, unfair advantage

What: What you know, whom you can reach or what you can build that most people trying the same idea cannot: domain experience, access to a community or a rare skill.
Why: Competing on effort alone is hard. An advantage in reaching customers is often worth more than an advantage in code.
Example: A student who worked in a pharmacy knows the stock problems and can get ten pharmacists on the phone next week.

Also called scratching your own itch

What: Building a product for a problem you have yourself, so you understand it deeply and can judge the product as its first user.
Why: It gives you a head start on understanding the problem and the motivation to finish. Many developer tools and templates start this way.

Watch out

Assuming that because you have the problem, many others do and will pay. You still need to check.

What: Public evidence that people want a solution: repeated complaints in forums, one-star reviews of existing tools, questions asked again and again, job posts for manual work, and people paying for clumsy alternatives.
Why: Signals let you test many ideas quickly before spending time on interviews, and they give you the exact words customers use.
Example: Twenty Reddit threads a month asking for a simpler way to split rent between flatmates is a signal; one viral tweet is not.

Also called keyword demand, search volume

What: How many people search for your problem, and whether that interest is growing, seasonal or fading, as shown by Google Trends, keyword tools and autocomplete.
Why: Search shows active demand in the customer's own words and hints at a future growth channel through search engines.

Watch out

Rejecting a niche business idea because it has little search volume; many good products are found through communities instead.

What: Existing products for the same problem show that people spend money on it. Your job is to find a group they serve badly, not to find an empty market.
Why: No competitors often means no demand. Reading competitors' reviews and pricing pages is free research on what customers value and hate.
Example: A big invoicing tool with reviews saying "too complex for a one-person business" points to a simpler product for solo freelancers.

Watch out

Giving up because a well-known product exists, or copying it feature for feature.

What: One or two sentences naming the customer, the problem, how they handle it today and what it costs them.
Why: Writing it down forces clarity and gives you something concrete to test in interviews and on a landing page.
Example: Solo tutors lose about two hours a week chasing parents for fees by WhatsApp and forget who has paid. Today they use a notebook and screenshots.

Go deeper