How do you validate an idea before building it?

Before writing code or spending money, test whether the problem is real. Here's how.

Cheap validation tactics:

• Talk to 10 potential users about their problem (not your solution). Listen for pain.
• Make a simple landing page describing the product; see if people sign up.
• Offer to solve the problem manually first (no product) and see if anyone pays or engages.

The goal is to be wrong cheaply and early. How did you validate (or fail to validate) your idea? Honest stories welcome.