A practical order of operations for finding and launching a startup idea

Executive overview

Most aspiring founders believe they need a great idea before they can start. This belief keeps them stuck — and it's wrong. The right approach is a three-step sequence: find a personal problem, brainstorm solutions with potential co-founders, then build and launch an MVP as fast as possible.

The goal is not to have the right idea upfront — it's to pick a real problem and iterate toward an answer.

Step 1: Find a problem worth solving

• Start with problems in your own life or work, not abstract market opportunities.
• Test any candidate problem on two dimensions: frequency and intensity.
• A problem that scores low on both is not worth building a company around.
• Aim for one of the top three problems in your life — those tend to be big enough to matter.
• "Meta" ideas (e.g. helping other startups) are usually a sign you haven't found a real problem yet.
• Work problems are equally valid: "I keep doing this manually, I'll automate it and sell it back to my industry."

Step 2: Brainstorm with potential co-founders

• Your first solution is a hypothesis, not an answer — expect it to be wrong.
• The point of early brainstorming is to find people you can riff well with, not to finalise the idea.
• A good brainstorm partner builds on ideas; a bad one shuts them down.
• Arriving at an idea together gives co-founders shared ownership and early commitment.
• Do not marry a solution — the problem is the anchor, not the product concept.

Step 3: Build and launch an MVP

• Ignore the voice telling you to incorporate, fundraise, or write a business plan first.
• Build the simplest possible V1 — even a spreadsheet-backed prototype counts.
• Get one or two customers before worrying about legal structure or funding.
• Early customers validate that the team can work together and raise the energy level.
• Traction also gives you leverage: investors respond to customers, not decks.

Two failure modes to avoid

• Pitching investors before building anything. A deck alone is no longer enough. Some founders can raise on an idea; most should be good rather than lucky.
• Outsourcing development without a technical co-founder. Outsourcers produce a V1 that requires constant iteration, which burns money fast. Founders often don't see the problem until the money is gone and they still have no working product. Getting a technical co-founder is usually cheaper and faster than outsourcing — and it just takes time and effort to find one.