How to pick the right idea?

I know an old post...but a framework that's worked for me: score each idea on three things and let the numbers break the tie, because your gut will just pick whatever you thought of most recently. First, how specific the buyer is (can you name them by role or situation). Second, whether one person can ship a usable v1 in weeks, not months. Third, how quickly someone could realistically pay you.

The buyer-specificity question is the one that's saved me the most time. "SaaS for teams" is a red flag. "A scheduling tool for solo bookkeepers at tax season" is something you can go validate this week. Whatever you pick, resist starting two at once — ten ideas usually turns into zero shipped, not because the ideas are bad but because attention is the scarce input.

10 products in 10 months 👍

Framework is trial and error.

Instead of picking one, I'd try to sort them through some sort of metrics that I think are important;

• How hard to market it, do I know any good channels.
• How long it'd take, will I get bored in the middle of it.
• How's the competition, how much they charge for similar service, how many users I need to call it a success, is there 100x more people that can I reach.
  ...

So far my favourite way to pick an idea is by checking on how much the problem is being mentioned in communities, forums, groups.

Usually people don’t mention what they want to buy to help them. I will see which problem related to my idea is most common. Chances are that could be the one most in demand, then reach out to them to test if the problem is bad enough for them to want my solution.