It's actually not that hard to get good clients on Guru and the sites you mention. You just need to differentiate from that Russian kid on something other than price.

Back in my eLance days, I charged $75/hour for my time and wrote good thoughtful proposals for jobs that sounded interesting. Those two things set me so far apart from the rest of the field that potential clients would essentially have two piles of proposals on their desk:

- Pile A: 100 broken english canned proposals quoting $14/hr and dripping with flakiness

- Pile B: 1 proposal from the expensive guy who sounds like he knows what he's doing

So the thought process then changes to "do I take a risk, or do I spend the money to do the job right". I'd only hear back from 1 in 10 proposals, but the conversation was always the same from there:

"Wow, you really nailed what we're looking for, but, well, you're a bit expensive. Any chance you can give us a break on price?"

"No."

"OK, well we've talked it over and we think we'd like to give it a shot."

Why not "charge" $90 per hour and then offer to "come down" to $75?

Why open the door to negotiation? If you're in a position to charge $75/hr with confidence, then charge $75/hour. With confidence.

I think you missed my point. Charging "$90 per hour," some people will just give you that; that's obviously better for you. There will still be people who hope to get a discount, and now $75 seems like you're doing them a special favor.

I'm curious about this as well. I've posted my skills on craigslist and sent stuff out to my friends, but I always receive requests from people who can only pay very little. Either that, or I'm not doing a very good job of advertising myself - which is a definite possibility.

From personal and professional networks. Work at BigCo or Startup or wherever for a while, make connections with real people, don't sabotage your reputation with shoddy work or burned bridges, then transition to contract work for the people you know, advertise yourself so you can freelance work from people you don't, self-incorporate/brand yourself/etc. It can't be done overnight, but very little worth doing can.

Definitely stay away from those sites unless you plan on doing volume/mass work to compensate for the low wages (or unless you have a big reputation). Like I mentioned in another post, at least 85% of my freelance gigs come from my existing network and their word-of-mouth referrals. I've found clients are much willing to pay more when they trust the consultant and a verbal referral from a friend can go a long way in providing that trust.

Some tips for this:

* Make sure your friends know what you do - update your Facebook profile, hand out business cards, start a blog, etc.

* Make sure your friends think/know that you are the best developer that they know and that you are the first person they think of when someone says "I'm going to need a website"

* Use LinkedIn - surprisingly enough, I've had old contacts connect with me via LinkedIn, view my profile and hire me for a project

* Exploit whatever niche you're a part of - if you are active in your church, approach them or another local church to see if they need a site. Apply same strategy for a youth group, social club, etc.

I'm in the Midwest US and I've used RentACoder (now vWorker) but you have to be careful. The key is to cherry pick specialty work that you can do quickly and still charge a reasonable fee for. In my case, I do almost exclusively hardware designs that I can knock out in a few minutes. Simple stuff, but I rarely get more than $250 or so. There is more involved work available, but I refuse to work for $10/hour!

I've bid on only three projects in the past year and been outbid on 2, 1 is pending.

From a network of people who work in real businesses who know what you do and that you want to book gigs (and subtly, that you'll take care of them in a meaningful way if they bring you work that you book)

Reminds me how once a customer came to me and needed his elance-made-in-russia-iphone-app fixed because it "mysteriously" crashed since apple "broke it" with the new iOS version. (ASAP!)

I took a look at the code and ... you know what happens in japan with that nuclear plant right now? Yeah, the app was the fukushima of code. It only hold together because of luck. Luck didn't last till iOS 4 where Apple seemingly changed something about how cocoa touch retains objects internally and the app exploded. Objects were randomly released and the programmers didn't care about any conventions. They returned objects with retain counts of 1 from a static lib and then released them somewhere later in another static lib. Not a single autorelease was used.

At least there was extensive doxygen generated documentation ... in russian (no joke).

You get what you pay for. And yes you can't compete with the people offering their services on odesk/elance. And you don't want to work for cheap asses that post "need augmented reality app for max $200" jobs.

I'm sure you could have eloquently described the code without comparing it to Fukushima. It is an ongoing crisis, part of a larger natural disaster, with many people still actively grieving.

It is a ridiculous analogy anyway. Fukushima failed, yes, but only after a magnitude 9 earthquake, followed by a devastating tsunami, overwhelmed its myriad safety systems. I don't know how any right-minded person could ever form the impression it was only "held together by luck".