I've been using oDesk exclusevly for over 6 years but i find that the quality of jobs and overall experience is on decline. There is a growing issue with spamming/phishing (including the glorious Nigerian money scam) on oDesk. But that's somewhat understandable and I'm sure moderators are doing their best to combat it.
However, there's a huge problem with the search algorythms on the site. In order to gain client's trust, odesk needs to make sure they display their "most successfull" providers at the very top. This can be done using a combination of the overall score/rating, hours worked, overall earnings (which is a direct reflection of the contractor's success rate), number of items in portfolio, tests taken, and so on. At the moment, if you search for ANY type of contractor you'll get VERY mixed results and honestly, I would not hire 90% of the first page of contractors.
I've been approached MANY times by clients that have already paid other contractors for a job that did not work out. This hurts odesk in 2 ways; for one you've linked a bad provider to a client that will think twice about outsourcing work again. Second, the next candidate/provider will have a much smaller budget to work with IF the client decides to repost the same contract. Either way you slice it, nobody wins.
What are your thoughts/ideas or opinions on the matter?
Sounds good in theory, but the problem is that sorting search results strictly by work experience will lead to a 'rich getting richer' cycle. As that happens, the 2nd and lower-tier of freelancers will start to leave the site for lack of getting enough work. And as the freelancer pool decreases, it empoverishes the ecology and will lead to a decrease in clients as well. After all, even qualified freelancers, when they join the site, start with nil work history and ratings. oDesk has to encourage clients to spread the work around, Of course, the shuffled results should not present irrelevant or low rated freelancers. They should do a better job of mixing the pool. But the best to worst ranking suggested isn't healthy in the overall scheme.
I am somewhat pleased to see the comment of Gyan and the discussion topic.
I am going to talk from my experience. I have found expert freelancers who charged the hell out of a new client. I have also found experienced clients who have taken the most out of a skilled new freelancer at a very low cost.
But this is always going to be the hard reality of Odesk. One can just try harder and harder and eventually most of the freelancers are going to find a suitable client whom they feel comfortable to work with.
But the site has recently got diseased with ridiculous job posting. I had shared my thoughts recently: **edited for Community Guidelines**
It's not just a matter of getting 'some' job but of spreading the work around. The competent freelancer who joined 6 months ago and has only, say, 200 hours is no less deserving of consideration than a comptent freelancer who joined 3 years ago and thus has a history of 2000 hours. But your suggestion would promote only the latter to all clients who search with the former appearing on page 10 or later of the search results.
So oDesk should apply a basic filter in terms of minimum feedback rating obtained in the category in which the client searched*** and then randomly shuffle the candidates. Sounds fair to me.
***One flaw of presenting the hours count and overall rating is that fixed price jobs aren't factored in (other sites mention total no. of jobs) and that the rating reflects all jobs. A client posting a Wordpress job isn't much interested in the ratings obtained by a Wordpress developer in some web research jobs he did on the side.