The first thing you should do is write it down. Write all of them down, in a notebook dedicated to nothing but your Eureka moments. If you don't have the notebo

The first thing you should do is write it down. Write all of them down, in a notebook dedicated to nothing but your Eureka moments. If you don't have the notebook handy, write it down on whatever is, and put it in the notebook later. If anything else related to that idea comes to you immediately(connections to news items you've read, statistics you've seen, specific features or UI choices you'd want to implement), write those down too, until they stop coming. Then, let it sit for about 3 days. Do your usual tech reading during that time. Does the idea keep nagging at you? Have you been writing down more related ideas on scraps of paper? Are you seeing connections to it in what you read, what you hear, what you say? If yes, then you might be onto something. So... 1) Take it to your braintrust. If you don't have a group of good, trustworthy friends who are also interested in various tech subjects, you should. Tell them that you've had this idea, and you're wondering what they think. Ask if they've heard of anything like it before. (A lot of times people will say they think X is like your idea. Don't be discouraged; it's human nature to take things we're unfamiliar with and look for something we can relate to. Whatever they mention might not be anything like your idea, but research it, and note the differences.) 2) Figure out what space your idea is in. Online video? Peer-to-peer? Social networks? Figure out if your idea is an improvement on current offerings, or something with a unique life of its own. If it's just an improvement(and it's not patentable), it's probably not worth starting a company for just to get it copied by all of your entrenched competitors. Instead, this might be a good opportunity to make connections in that space by offering your idea to whichever existing service you like the most. If it is unique, then move on to... 3) Evaluate your commitment. It should be about 9 days after you had your idea by now, or at least a week. Does it still excite you? Really excite you? If there appears to be a market opportunity for it, how much effort will it take to enter that market? Are you willing to make that effort? What if it took twice as long as you think it will? 4) Start the startup process. We have plenty of "A-ha!" moments. If you've made it through the first three steps with the right answers, you've sifted out a moment that has the potential to be more than just an incremental improvement on something that already exists, and something that you yourself can commit to. You still need to do more research, such as "Why hasn't anyone else thought of this before?" and "Are other people going to want this improvement?". The good news is that if you can find a person who's as excited about it as you are, you may have found a good co-founder. And that's the best validation of your idea you can have at an early stage.