99designs is an internet design "crowdsourcing" platform for hosting design contests. You post a design brief about what you want designed, choose a prize (from 99 USD and up, depending on what you want designed), and designers from all over the world post their complete design entries. Some weeks ago I ended my first logo-browsing marathon there. While I did get some valuable tip from a couple of articles (exratione.com, quixey.com) that I read before starting the contest, I had unanswered questions that I now can answer and share with you:

• The prize you set doesn't only affect the number and quality of design entries, it also affects the time you have to spend going through amateurish, silly entries. I spent 310 USD which is just a little more than the minimum, to be listed higher, and several of the designers who entered submitted entries that I simply eliminated on sight. If you spend more, hopefully you'll spend more time evaluating good quality, creative creations, although the amount of bad entries might rise too.
• You get new entries and designers by making a good brief, having a high prize, and most importantly inviting designers. I spent a lot of time writing individually to the ones I invited, then I gave that up because it took too much time. I invited 57, and 13 of them gave entries, but others have invited 200. It seems the more you invite the more designers will join, but don't invite unnecessarily, then you'll waste both yours and the designers time.
• The best way to get new entries is to give feedback, especially individual feedback. Simply tell the designer what you like and don't like and what you'd like to see improved. I was very conservative with four and five star rankings, keeping the good designers in the loop.
• If you know what kind of font you like, then specify, if you know what you want to use your logo for then specify and ask specifically for the designers to provide those uses, f. ex. provide thumbnail (for Twitter et. al.), favicon (for browser), black and white version, and different use of the logo colors. I mentioned favicon and twitter, but out of the 282 entries, only one included favicon. This means that I don't see if the logo is suitable for favicon use, and unless the winning designer is kind, I'll have to acquire or make the favicon separately.
• Don't start your contest on a weekend. The last couple of days will probably in any case be the most active ones, but you risk having a very slow start if you start on a weekend. I started mine on a Friday just before noon, got the first entry after 1,5 day and the second entry after 3 days. Next time I'll try to start on a monday afternoon.
• If you are active giving feedback and inviting designers, I do not believe paying for being highlighted is worth it. Better spend those money on the designers.
• Guarantee your payment when you get the opportunity.
• Blind or not blind: you might think that making the contest blind is mostly for the designers to feel more comfortable because they don't have to worry about being copied. What I experienced with my open (non-blind) contest is that the designers seemed to copy each other too much. A lot of the entries used colors and ideas that I myself had only rated with three stars and didn't really like that much. Such "un-creativity" would probably be avoided with a blind contest, but on the other hand, the designers also used elements from each other that improved their own work. Such elements can also be steered with feedback, however.
• You can look for color inspiration at Cssmania and Colourlovers and specify them in your brief or comments.

I've also suggested some improvements to 99designs, that you can vote on at uservoice. And lastly: here's the logo that won the competition: