Differentiate on Quality, Never on Price
I would write proposals by hand, in literate English, after reading and considering the project description. That alone would differentiate me from 90% of applicants. I'd then include a ballpark estimate along with my hourly rate.
The response I got back was pretty much always a variation of "Wow, you really understand our needs a lot better than anybody else who's responded. Unfortunately, your bid is 3 times that of the next highest bidder. Any chance you could come down a bit on price?"
My response to that would be some variation of "No.", and their response to that would generally be "OK, let's give it a shot."
Here's what's happening in that situation. I wanted the project owner to end up with two piles of bids on his desk. Pile A would have 150 English-as-a-second-language generic bids that didn't address the project at all and quoted $7-$14/hr rates. Pile B would have exactly one bid from a guy who obviously knew what he was doing and would definitely deliver, yet was quoting a much higher price.
At that point, the project owner has two choices: Sift through all the garbage in Pile A in hopes of randomly finding somebody good. Or spend a realistic amount of money with the guy from Pile B.
I got plenty of work this way, and as an added bonus it did a good job of filtering out the unrealistic "I need an eBay clone for $400" types.
Everybody says this and it's true: to get started you have to make a couple of bids at sweatshop rates in order to build reputation. I found explaining this reason as part of my initial bids to be a successful strategy. Your first week (or longer!) you'll earn a lot of nothing. Once you've gone through this probation period, though, you'll start earning decently.
From what I can tell, there are plenty of users who reply to a bid with "yes, I can do this, I have done lots of projects like this, CHOOSE ME!" So, instead, read each request fully, and ask lots of questions. I've found that asking questions first and only entering a bid amount after receiving answers to my initial questions to also be a winning method.
Despite plenty of articles to the contrary, it is more than possible to earn a decent living through sites like this. By decent living I mean $5-$6k/month, your mileage may vary. To do so, you have to have the balls to bid high - with justifications - and accept that you're not going to win very many bids, but that the ones you do win will more than make up for it.
Finally - and I've again read plenty of articles that contradict this - if you're competent and do a decent job, you will get repeat work. This has happened to me several times, and for the last few months I've been turning work away.
I also posted a few projects at elance and got some real bids from locals, but the off shore firms almost always sent in form letter type bids. They weren't appealing at all.
If you are going to do it, I suggest finding something you do want done and are willing to pay $100-300 for. Post it and see who your competition is going to be and how much work you'll have to do.
If you are good with JavaScript, HTML, and CSS, I have work I'll pay you $100-300 to do. Do you have a portfolio? It may be good to post it just so we get an idea of where you fit skill wise. Maybe you could peruse craiglist once in a while. I've posted links to sites there and off shore people email me and are like, "Hey, I could redesign your website!" So I said ok and they did a good job.
No experience using the other sites. I did look at a couple sites like that back then, but the projects weren't a good fit and the values were too low to justify the customer acquisition costs in terms of time. If it takes 5 hours to find a project and it's only paying $300, $60/hr is the most you'll make if it takes you zero hours to complete the task. If it takes ten, you're down to $20/hr. It'll be hard to keep that ball rolling unless you work your way up to bigger and more valuable projects.
I make roughly $250-$300 a week in my spare time on there (roughly 10 hours?).
What do you think of the assignment that seems to offer peanuts for complex task?
I ask because I see lots of iPhone stuff that seems to require a lot of work and paying around $150-$300 in oDesk.
I'm not sure I understand your second sentence, but part of making some decent money on oDesk is quickly weeding through and finding the jobs that want a skilled worker and are willing to pay for him. There's plenty of information given with each listing and I careful select what to apply to so I'm given the majority of jobs I apply for.
There's iPhone work available for $x,xxx if you really know your stuff. There's always buyers trying to "clone facebook for $10" etc.
If you take as many tests as you can and completely fill out your profile, it makes a world of difference in selling yourself. I'd also suggest pricing yourself a bit higher if you think you're good. I imagine you shouldn't have any problems making $300+/mo off it
The best way to land jobs is to be the best person for the job. Some buyers are able to cut through the fluff and identify the quality. These buyers aren't nearly as concerned about price.
The way to be the best man for the job is to have a strong profile in the area you are bidding for. For example, if you are bidding for Drupal jobs, then be a contributor and /or have some useful add-ons under your belt. Express confidence that you are the right person in your proposal and back that up with your credentials.
Speaking of building add-ons, while using this method to build up your profile, consider selling something if it gets a lot of interest. You might make as much from selling add-ons as your target for extra monthly income.
In certain jobs you may even outline how you would do the job and what tools you would use. This works great for smaller jobs because it educates buyer on what you will be doing before you even get started. This really sets yourself apart from the generic proposals.
There are certain jobs which might look decent but you can clearly see the buyer is going for cheap in an area that lower cost providers should be able to easily handle. If you get this vibe and your rate needs to be higher then move on. Usually these types have a low hourly rate posted.
Look for jobs with a fixed cost. You can often beat your target hourly rate if you find something you can do fast.
Done right, all that competition won't matter much.
I always tell me clients that I have a 8/9-hour day job. I update the client every few days through email. Sometimes I work at night and sometimes I don't. When I work at night, I spend around 4 hours maximum -- average is 2 hours. On the weekends, 6 hours. I was still able to go out with friends and have fun. Mostly TV time and video games were cut-off - except for a brief 2 month re-play of Diablo II :P
I do mainly web development / web design.
Right now, I am trying out a new career that doesn't involve the internet. But still though, on my free time I work on pet projects I have listed down a few years back.
For 2 years, I used it as one source of income for my freelancing business. But like I said, I didn't compete on price and only bid for the interesting projects and serious clients.
1. Differentiate on Quality, Never on Price
2. Find a niche, fill the niche
3. Avoid cheapskates
4. It works, be persistent