There may be an average for your particular niche, however I believe it varies drastically on the whole. Largely depending on the work you do, experience, and your contacts. Also, I believe the rate you charge gets you completely different clients. Charging $50 an hour? You will get people who don't pay, don't know what they are doing, and are hard to work with. Don't change anything except your rate (e.g. $150 an hour) and you will see completely different clients.
I personally work only off retainer at a rate of ~2k per day. I think for my line of work this is on the lower end. Retainer must be paid upfront. I have fewer, but longer lasting clients and projects. I am VERY good to them. I believe I approach the freelancing game different than most and it nets a lot of referrals. I often turn down work and I am a stickler on who/what I work on.
NOTE: 90% of the time I work for myself solo and 10% of the time I freelance, mostly out of the desire to get new perspectives and work with others.
Someone coming in cold to freelancing will spend time building up to this level.
They may need to take $50/hr jobs for a few months to build up a portfolio and references, and gradually raise your rates. Good, reputable clients will stick with who they know before hiring someone from the 'outside'. After that, referrals from their current contractors, then friends at other companies. You'll be working at this tertiary level for a while when you start. It sucks, but you just have to put in your time.
As a general rule of thumb, I also do not recommend taking gigs with individuals. The line can be blurry, but generally 'guy with an app idea' is not anyone you want to work with.
Another method you might consider is milestone based payment. You have a project-as-a-whole cost but have set milestones and when they are met you get paid. This is nice for both parties at the beginning of a relationship to "test the waters". If they flake out on you, you aren't out as much time. They also have less risk.
You can get creative with how it is broken up as well. I've taken the approach before of having the first few milestones be less expensive to get a foothold and get them "hooked" on my skills. At the same time, I am evaluating them on communication, requirements, and how we work together.
FYI: Retainer based pay is usually a hard sell to make but once they do it, they love it. I have yet to have a client not want to do it after we have started down the path. However, I have had several potentials deny me outright. Since I do freelancing part time this is okay by me. I really favor long term/large clients. The hustle of getting clients is annoying to me. I'd rather give a decent discount (e.g. 35k vs 60k) for a client who knows they will use a whole month of my time (all I do per year). When I did the cost analysis, it was about even. However, I'd rather be coding than marketing another client.
I second the milestones idea. I used to do this back when I was freelance web developing and it helps the client feel more comfortable knowing they're paying as they see progress. (Helps you a ton too cuz you get money as you go along).
This was one of the major features of my new startup matchist (http://matchist.com/talent). You create milestones for your client and they prefund each milestone before you start working on it. Once you're done, we release the payment to you.
I personally work only off retainer at a rate of ~2k per day. I think for my line of work this is on the lower end. Retainer must be paid upfront. I have fewer, but longer lasting clients and projects. I am VERY good to them. I believe I approach the freelancing game different than most and it nets a lot of referrals. I often turn down work and I am a stickler on who/what I work on.
NOTE: 90% of the time I work for myself solo and 10% of the time I freelance, mostly out of the desire to get new perspectives and work with others.