Most founders start from the wrong end. They fall in love with a solution and then hunt for a problem to attach it to. Better ideas come from the opposite direction: notice a problem worth solving first, then build the smallest thing that solves it. As Paul Graham writes in How to Get Startup Ideas, the trick is not to think up startup ideas but to look for problems, ideally ones you have yourself.

Notice, don't brainstorm

The best ideas rarely arrive in a brainstorm. They show up as friction in your own life and work: the thing you built a spreadsheet to cope with, the task you dread, the tool you keep wishing existed. Graham's warning is to avoid made-up, "sitcom" startup ideas and to trust the organic ones instead. Keep a running note of your real annoyances for a month and the patterns start to show.

Scratch your own itch

Building for yourself is an unfair advantage. You are the customer, so you already understand the problem, the jargon and where the existing tools fall short. 37signals built Basecamp to manage their own client projects, and Pieter Levels' Nomad List grew out of a spreadsheet he kept for his own travels. Neither began as a grand plan.

Sell pickaxes

Look for places where money already moves. People and businesses happily pay to make money, save money, save time or reduce risk. The cliché has real roots: in the California gold rush, Samuel Brannan got rich selling shovels and pans to prospectors rather than digging himself. An idea that sits next to an existing budget is far easier to monetise than one that asks people to form a brand-new habit.

Filters for a real idea

  • Is the pain frequent and urgent? Vitamins are nice. Painkillers get bought.
  • Can you reach the buyers? A great idea for an audience you can't access is somebody else's idea.
  • Will they pay? "That's cool" is not a business. Someone reaching for their card is.
  • Can a tiny team ship a v1? If it needs ten engineers and a year, it is not a bootstrap.
You don't need a brilliant, original idea. You need a real problem, a reachable audience and a willingness to start small. Originality is overrated. Traction is not.

Got a candidate? Don't build it yet. Validate it first. For a steady stream of structured opportunities, Trends.vc is a good prompt.