It seems like every time I'm on a forum or read a well-trafficked blog about internet marketing, there are people who are asking how they get ideas for a niche. They seem to be stuck trying to come up with ideas for topics/markets to build their websites around.

I can't understand this. There are ideas everywhere. My problem isn't coming up with ideas for niches, it's trying to stay focused on the ones I'm already working on when I get another 10 new ideas practically every day.

I think a lot of these people aren't really stuck for ideas for niches. What they're really asking is how to come up with ideas for successful niches. They want to know how to identify the markets that are going to make them money.

You can never be entirely sure until you actually do something to test the market, but there are some things you can do to improve your odds.

1. Check for magazines in the market you're considering. If there are magazines being printed, you can bet people are buying stuff in that market since the magazines' success depends on advertising. Companies don't pay for magazine advertising if it isn't converting into sales.
2. Look for a good selection of products on Amazon.com. If there are a lot of books on the topic, people are interested in it and buying the books. Or if it's a different kind of product, having a good selection to choose from means it's more than likely selling.
3. Check eBay for auctions in that market. If there are lots of auctions going, and a decent number of them being won (say 40% or more), it's a market with buyers.
4. If it's an information based market, check Clickbank for products. You can tell from the Clickbank stats whether it's selling, and if there are several products on the topic, chances are it is.

In the end though, the only way to know if a market is going to be successful is to test the waters. The best way to do that is with some pay-per-click advertising. With a minimal ad spend, you can tell whether there is traffic and whether or not they're action takers.

Send the traffic to an affiliate offer or even to a squeeze page to get them on a list. At least you'll know if they'll take action that way, but you need to send them to an offer that requires some kind of payment if you really want to be sure they're going to actually buy something.

It's much better to spend a few dollars on Adwords to check whether a market is going to be worth pursuing than spending several hours or days getting a website set up and working to send some free traffic to it, only to find out it's not going to work.

Do it enough times, and you'll find some winners that will more than repay any losses you take on the losers - I can guarantee it.