If you're just starting out, pick the one you can finish. Most people never follow through, and it's a pretty basic lesson to learn.
Once you got that down, pick ones that excite you.
Later on, you'll think about stuff like potential market opportunity. And whether people actually want what you've Built. But I think it's important to learn HOW to build first. If you can finish things, you're already well ahead of the pack.
Good advice. Beyond that, I recommend surrounding yourself with critical, intelligent, trustworthy people. I bounce all my ideas off a core group of friends, and generally don't bother with anything until they all give the green light. Besides that, it's just ROI/risk evaluation. Unless you're already a millionaire, it is best to focus on lowest possible investment for the highest likely return. Later it may be worth taking bigger risks.
Once you have started a project, you might feel this idea is not worth doing and you will have some other ideas. But don't change your mind. Finish what you are already building. If you leave this project incomplete and switch to another, you will never have a complete project. Instead note down all the ideas you have. Once you are done with current project, go to that list, see what you want to do next.
2. Do I need to teach people that this is important...or do I get to exploit existing demand?
3. Is there some sort of barrier to entry?
4. How long will it take to build a minimum viable product?
5. What the competitors are like, their current reputation, their current SEO rankings, their current adwords campaigns etc.
6. Then I usually build a quick fake page to see if there is any demand. I'd rather piss off 200-300 people and see what the actual demand is like, instead of relying on them to tell me what they think they want.
7. If everything works out, I get to work. If not, I get to save 3 months chasing an idea that has no potential.
There's a story about someone telling Einstein that he keeps a notebook on hand to jot down all the good ideas he's had. Einstein said he wouldn't need such a notebook. Because he has such a great memory? No, because he'd only had one or two good ideas in his life.
Seriously, if none of them leap at you, grab you by the neck, and insist on existing, it means you don't have any good ideas.
But this doesn't mean you should be discouraged, it just means you should be in "getting good ideas" mode. And the best way to do this is to get out there, experiencing problems, solving them, making mistakes, correctly them, feeling stupid, making progress, all the while trying to understand things, wondering about things, being exposed to lots of different ideas from different places, and trying out the ideas you've had, if they interest you, just out of curiosity. Get busy living or get busy dying in Morgan Freeman's voice. "Evaluating ideas" is a way of dying, IMHO. At least, that's how it's been for me when I've got stuck in that way of thinking, anyway.
Then one day, you're looking at something, and you think "you know, everyone is doing this in a stupid way. why don't they do it this way? it would be so much better." You whip up the simplest prototype you can think of, in a couple of hours, just to check for yourself that the basic idea does basically kinda work. Then you whip up something, in a day, that others could try (a day's work: that's how incomplete and limited and trivial that prototype should be - an MVP), to see if anyone else is the slightest bit interested - if they are, that gives you the information and the encouragement/fuel to continue. And suddenly, without quite meaning to, your journey has begun.
That's what happened to me, anyway.
Take the heaviness away. That corporate/academic/executive approach of "criteria" and "evaluating ideas" doesn't make giant corporations agile - and it's only going to drag you down. Insight and intuition come not from methodologies, but in spite of them; they come from contact with reality, and from you. And insight and intuition is where the great startup ideas come from.
disclaimer I'm not a wildly successful entrepreneur, just a successful-enough one. A cool idea from many years ago is still making money for me today. This is all just IMHO, my experience, YMMV.
2.just won't get out of my head even after it's been days since I first thought of them
3. keep luring me to go back to work on them even after the initial honeymoon period is over and I've been through days when I felt utterly bored of them.
I have only one suggestion which helped me pick out the good ideas from the bad - to get external opinion. It's almost like humans have a self defense mechanism by which they can't adequately find flaws in their own selves and ideas. Thus, your idea might be the biggest thing in your mind, while other people might feel otherwise.
That's why I told everyone about my idea - presented it in startup events, showed people demos, put it up on public mailing lists and wikis. The idea got battered and beaten down and each time there were little changes. There's this huge hype about being a "stealth startup" and a fear that people in some web shop will steal your idea and clone it.
tl;dr - the idea that you start with is the one that people can't completely tear apart and obliterate.
Also, the idea is pretty much useless. If I had an idea that I'll make a touch screen phone with an app store, nobody would've taken me seriously. If I'd pulled an apple and built one, people would've. Three months into full-time development of a product, I can only say that execution is pretty much everything.
Summing up what's being said in the comments, the two main points are:
1. Subjective: are you really really willing to spend a lot of time and energy (not to mention money) on the idea, to the point that it becomes the only subject of your thoughts? (Paul Graham wrote two inspiring pieces on this: http://paulgraham.com/top.html and http://paulgraham.com/determination.html)
2. Objective: is there really a market, is people really willing to pay for your solution? etc... There was a submission a while ago on HN about a technique dubbed "Ghetto Testing" that was quite interesting: http://news.ycombinator.com/item?id=1485057
I'm really guilty of spreading myself too thin with projects. Recently I asked myself a different set of questions to evaluate what projects I wanted to work on, and it made deciding very easy:
1) Which ideas can be turned into a company?
2) Which have the most attractive success scenario?
I spent too much time focusing on any idea that could make money. Well, the problem is that making money really isn't that hard if you know how to ask for it. Every idea that came into my head was good enough to work on - if you ask yourself question number 1 a lot, you'll have the same problem.
Once I started focusing on question 2 everything became much more clear. You have to ask yourself a lot of questions to get to an answer; the questions (and answers) are far more personal than you can ask advice about. If you're going to give up a significant portion of your time to succeed, it's important that there's a big enough payoff for you at the other side. For some people, this is purely financial; some people want to make "popular" software that everyone uses. There are people who just like working on interesting problems, and people who really just want a day job where they can be "the boss".
If you decide what's most important to you, it should be much easier to decide what to work on. You'll end up with the added benefit of being much more motivated to work towards a goal that is desirable to you; not one that everyone else thinks is desirable.
I have ideas all the time. A few good ones and a swarm of bad ones. Filtering them down is sometimes tough.
I believe you should work on the one you feel the most passionate about. Ideas do not need to be great to succeed. But, unless you have a fantastic game changer of an idea, you need to work hard and passionately to get it off the ground. The more you are vested in the project emotionally the more likely you are to always push through.
Hate cats but have a great idea for a new cat toy? Screw it. Love dogs but have an iffy idea for a new product (doggles anyone?)? Go for it.
As Edison said, "Genius is one percent inspiration, ninety-nine percent perspiration." Pick the project you are most willing to perspire over.
Careful with subjective measures. Measure interest.
Invest some $$ in google ads, have a purty-looking landing page that says "coming soon!" Build a site and ads for each of your ideas and chase the one that has most interest.
The one I would want to build even if I can't earn anything with it and could only use it myself. Once that one was determined to be commercially viable.
In your case I think you are should learn what are your strengths and weakness. So you should ask yourself what type of idea can I develop that can give me some feedback about my capacity to built something valuable. Once you know something about yourself you should look for a good team. A good team is the one that gives you the best opportunity to built the valuable idea that you have.
* Take out a spreadsheet, put down the names of your ideas in every row, put your evaluation criteria in the columns (Ease of development, New Skills, Current Progress, Real Life Value, etc).
* Add a row at the bottom to specify the weights of your criteria.
* Add a column to sum up the scores.
* Sort Z-A by total score and pick the top 2/3.
There's a saying that "you are married to your business" (or something like that). So I normally go beyond the tangible things to pick ideas (or spouse)...
I pick the one that I keep dreaming of and can't stop thinking about every seconds of the day.
And it works alright so far (both idea and spouse).
I tend to go with magnitude of possibilities. Specifically, if I'm still thinking of new applications/markets for the idea days later, it's probably worth pursuing.
Sounds good, but actually imo flawed advice. Almost every new idea I've had that has "kept me up at night" has started to seem much more mundane after a few weeks of work.
I have a list of ideas and I rank the ideas by how much I like them. So the top idea in the list is the idea that I'm most passionate about.
Then from the top of the list I think about these factors:
1) How big is this idea, can I achieve it with the time/resources I have budgeted for? Resources can be funding, helpers, co-founders, technology, etc
2) What is the market size of the idea? Can my team support the overall vision?
3) Who are the competitors? Or is there a competitor? If there is a competitor, can I compete with the available resource/time that I have?
Start from the top idea and if any of the questions raises a red flag, move on to the next idea. You don't have to use the 3 questions I provided above, just add your own questions but the methodology is the same.
Once you got that down, pick ones that excite you.
Later on, you'll think about stuff like potential market opportunity. And whether people actually want what you've Built. But I think it's important to learn HOW to build first. If you can finish things, you're already well ahead of the pack.