Does anyone else think the way freelance work is allocated is ripe for change?
Seeing the positive response to sites like http://tinyproj.com and http://weekendhacker.net, it seems like highly competent freelancers would benefit from a middle ground between the somewhat impersonal market of eLance/oDesk and the inefficient method of seeking work via personal networks. Something like a "hackers guild" of people who have been vetted for quality work. Projects could be semi-automatically assigned to single or multiple individuals based on the importance of the project as measured by the dollar value attached to it and the skill of the individuals as measured by their historical feedback ratings.
Maybe this is just my naive impression as an undergrad approaching graduation who would prefer freelancing to a 9-5 or even a startup but hasn't come to terms yet with the fact that inefficiency in the matching process and initially low rates are unavoidable. If not and you'd be interested in building or using such a system, I'd like to talk to you. It also relates to emerging things like the "reputation graph" and transitive delegation that can enable these sorts of systems, and are going to be huge (see http://blog.hyperarchy.com, http://liquidfeedback.org, http://rglabsinc.com).
But there are a few of these coming from HN readers; they just never pick up momentum. The thing is that MOST 'employers' also like to pay $500 for something which 'we know' cannot be done under $20k.
I did a lost of tests and spending 10x$500 for a project I know is worth $20k actually works very well. You just put 10 bidders separately on the same project and 1 or 2 turn out great. While the rest is of horrible 'quality', if they deliver anything at all. Yes you have to 'sit on them' more than you probably would for $20k teams, but it saves $15k. And i'm not kidding here.
This was actually just testing; I have regular people to do projects who I know and trust (and pay more obviously). We were trying to get ideas to open our own freelancer site; we gave up in the end because for employers there is not much upside. You have heard it before and you'll hear it a lot of times; there are complete novices who get nice stuff done on these sites without paying much and not caring about quality, code as long as it works. And that's on freelancer sites, sites like Digitalpoint forums are much much worse; you can get complete designs done for $5-10 and have projects implemented for $100-500 and that's considered a lot usually there.
We tried different prices, but they didn't seem to matter much. However we did not try with the same project, so it's not really the same case (we didn't have that much cash...).
We did see the same with admins and even quite high up (in the rankings on freelancer) ones; the same thing applied. We had an admin team which had super rankings, but their prices were insane, but we thought ok, let's try. So 4 admin teams, 8 servers, 2 each. 2 had ratings and 1 high ratings. 3 were actual teams, 1 was 1 guy (spoiler: the 1 guy won). They all were priced about $50-80/month per server for everything (yes... everything, including monitoring, 24/7 calamity resolve, etc).
During intake, the 2 rated teams performed very very well (spoiler: they do have 1 guy in their office who knows stuff who does only intakes all day, that's me presuming, but he), the other two less professional, but okish.
The first hick up was the highest rated company not knowing what an ssh key was. Or GPG (me asking for the public key for gpg to send over the ssh key must've sounded like alien speak to them)... They kept asking me for the cpanel root password. I usually don't run cpanel as I don't like it, nor does it run on debian.
This turned out to be the actual difference between the intake person and 'the rest' of the company (i demanded a total over these teams of 10 different admins) that they ONLY know cpanel. They absolutely did not know anything about linux but the sheer basics... Amazing.
The one person (coincidentally also the only one not from India) did know debian and was actually good, asking very low prices to get good ratings. We gave them and he's still managing 2 servers.
But how will ANY site trump this common trick for programmers OR admins. Yes using Odesk with camera works. But i'm not spending my time watching someone else, well actually a lot of someones, otherwise my scaling is limited to me watching someone, sit and write code. So usually the intake is with some person who is (very) good and when the cam goes off from the intake (or skype) you get some junior incapable of doing anything. How to prevent that and still enjoy the scalability of outsourcing? I mean if i have to check up on every person of my team, all day, I am better of putting someone next to me right?
There is so much room for improvement in this space.
I might be working soon on an awesome project that seeks to be the missing link for hiring in the 21st century. And, we have a way around the chicken and egg problem that dooms most attempts, im very excited about it.
I use oDesk to find Python/Django work. Although most of the work hasn't paid as well as gigs away from oDesk, it's worked well for me while I build up to properly paid gigs: I've just got my first one for years, after focusing mostly on moving to the US from the UK, living in an Ecovillage and raising my now-four-year-old. It took a while to get used to the oDesk tool taking a screenshot every few minutes - I found it quite humiliating at first - but after my wife pointed out it's no worse than having people glance over your shoulder in an office, I sucked it up. I like knowing I have access to the arbitration that oDesk offers if you use this tool. The automatic time tracking helped me overcome bad habits, too. I like the detailed time logs, and I like being paid weekly.
One thing that pisses me off about these sites is that there's so many of them. See, reputation is worth a lot, and it's not transferable between sites :( You work hard to gain a high score on, say, vWorker, but then at Odesk you're a fresh dev with 0 experience, unlikely to get any piece of work that can really bring normal pay.
Sure, but when just seeing the "name and rating" of bidders, this detail isn't directly visible. Many times buyers decide very quickly whom to give the work to, based on a very superficial glance on the ratings, without entering each dev page and examining it in detail. This is mostly true for the small jobs.
Thanks for posting this here, Derek. Definitely going to employ some of these techniques in our next round of hiring.
I especially like the idea of getting multiple developers working on the first milestone, in parallel. Sure, you pay a little more but it gives you a better chance at landing the perfect developer for the 'full enchilada.'
Seconded! When I saw the title of this post, I hoped more people would chime in about their experiences outsourcing. At this moment I used one contractor from odesk (for some audio recording) and it went very well (beginner's luck?). Hoping to outsource some artwork soon.
If I'm really desperate, I'd go back to vWorker, but I'd have to be very, very desperate.
I haven't tried out eLance or oDesk but I did check them out around the same time I used vWorker and found no real difference. It would'nt be worth the time I spent on the sign up process.
Tinyproj looked promising but by emails stopped sometime the other week. If anyone can do a Stack Overflow to these sites then I'd reconsider, but not until that day.
BTW: It would be awesome if someone could start a site with projects for BEGINNING Freelancers or for people with less experience than usual.
It could make sense for smaller projects and could even feature something like more experienced "buddys" who watch over the process? Just some ideas :)
I've learned programming on a website similar to elance (rentacoder, now known as vworker). Started with simple logo design jobs, advanced to simple php scripts, then c++, java, etc..
It was worth for me, coming from a developing country where $200 back then was an average month's salary, but i don't think it's too tempting for western countries.
For being successful reputation matters a lot. I was the top result for "android" queries on elance for over a year, and that allowed me to bid over the average, and that helps. I ended up eventually with a more permanent contract that pays better than most jobs on elance, but still not as good as a lot of Silicon Valley jobs.
So, that's where elance stands. Perfect if you're in a developing country :)
I'm currently contracted on it on an indefinite contract. I've also done a few small jobs with them and they can work to help make ends meet, but it can be hard sometimes to find something more than just small projects.
Aside from finding work, the software that oDesk uses for time tracking can be very annoying. The number of times that it's managed to decide that it can't upload the data to them recently has become a bit of a problem for the team I'm on.
I'm using oDesk for the past 3 years. It just worked fine and we are a 21 people team of developers now (from 2 people). Here's my profile http://bit.ly/vtv7Vx
Pros:
1) The hourly mode worked fine in my case. Get paid for the hours you work.
2) You don't need to send an invoice to the client. oDesk do it weekly and transfer funds to your account.
3) Good amount of money if you are living in a country like India.
4) 80% of time the employer will be an agency who outsource their work. If you are good you'll get their future projects.
Cons:
1) Most of the employers need cheap work.
2) Its difficult to find potential longterm clients and big projects.
3) You will face lots of employers who already lost money because of bad providers.
4) A single bad feedback from a bad client can make your profile rating down (the no. of bids you can make depends on the rating)
Now my team is working on 3 good startup projects through oDesk.
I think we found a programmer to write some Windows mouse hooks for our java app. Paid him $100 for probably an hour or more of his time, so I guess you can find good developers.
If you're cheap, you'll get cheap programmers, that's about as basic as it gets.
I used Elance and it worked pretty well. Be careful though. I posted project specs and had about 15 bidders. I found out once development was done, one of the bidders who didn't get the job liked the idea so much he wrote his own version, based on specs I provided for the bid. He even used the same name. So, on the appstore, I had to rename my app AND had a competitor immediately. Luckily, Apple and Elance took care of it and his product was removed.
I really liked the way elance gets you in touch with the developer and handles payment really well. Also, the communication system worked well. Would do it again.
It took me some time to find good offer with fresh account and get used to screenshots, but I liked it. Pay was decent and surveillance helped me to focus on work and separate it from breaks.
I've used oDesk to hire developers & SEO article writers. From a hiring standpoint, oDesk is nice because you can review screenshots of work hours submitted for payment. I've never run into an issue where the contractor was trying to get paid for time they weren't working on our stuff.
I haven't tried other freelance sites but I do remember researching a few before settling on oDesk for its payment controls.
If you're willing to work within the bounds of the system, I'd recommend giving oDesk a shot.
I hope you realise that judging programmer performance based on how busy they appear provides perverse incentives that are bound to lead to grossly inefficient behaviour.
This kind of system has been gamed for decades if not centuries in physical workplaces and odesk prides itself on perpetuating it in the virtual world.
For some types of work this is simply wasteful, but for software development it's outright destructive. You get below average programmers because they are benefiting most from the system, and you get unnecessary code that you have to maintain down the road.
I think odesk gives you a false sense of "value-for-money". But I guess it depends on what your alternatives are.
You only 'go to the video' when you feel you've not progressed as far as you'd projected during the particular sprint (taking into account the contractor's skill set & hours worked).
It's no different than a boss checking up on an in-house employee to see why their output is not optimal.
I do agree with your sentiments to a point though. I don't want a contractor that appears busy. I want one that is busy.
ODesk worked great for me, stick with very straight forward / easy tasks, I was paying $3/hr for someone to do data mining. I ended up with a great excel doc with over 3000 related sites & email addresses, very useful to my startups initial launch
The OP asked if you had used these services to find work. I think your response counts as a vote for "I've tried, but was unable to find a project that was worth my time".
Depends on which country you are in.
In India, a junior developer in an IT Services company (such as Infosys or TCS) makes 20-25k INR a month. Roughly about $3/hour.
While I would be hesitant to call it immoral, it can be alarming for a westerner to see that rate and realize the inevitability of it all. That being said markets find an equilibrium so long as they remain free, we have all kinds of things like unions to corporate collusion to corrupt governments, that work to keep them from actually being free in the real world. The reality is in some countries $3/hr is a respectable wage and corporations are exploiting wage and labor imbalances to pay that livable wage in one country while marketing their wares in a totally different economic environment. It may be immoral that we allow companies to exploit trade imbalances to the detriment of the people that are trapped in the reality of different economic conditions (e.g western workers), but that is a really complex subject to untangle.
To me, the part that I find perplexing is that if technical workers around the world knew exactly how short they are in supply they would understand their leverage in the market and not be asking for $3/hr. The thing is in some countries their is collusion to keep that realization from happening. I am not suggesting massive corporate conspiracies but rather people not knowing their true worth by gauging themselves and their income to local cultural norms instead of looking at global supply and demand and adjusting their rates accordingly. The reality is a guy in India or China could be charging $15 to $20 to $30 hr depending on specialty and still be competitive in the market.
That awakening happened among western technical workers and I think it will happen among other cultures, the cold hard reality is there is a small percentage of the population that choose technical career paths, and even fewer who are actually good at it. Once one realizes that the balances shifts in their favor.
I guess my point is in our field we only have our selves to blame if we are asking to little for our services. At one point outsourcing was used to leverage the price of technology down, but that has come back to bite the creator, it created new technical markets and constrained the supply of talent even further.
Millions of people would risk death for a chance at making 3usd an hour. People still die because they can't buy food. If westerners isolate themselves economically from the worlds poor in a misguided effort to remain morally pure, the worlds poor will suffer more, not less.
Elance works fine for Ruby work, and most of the time you can get better hourly rates than your typical full time job in the US. I think having a US flag in your profile also helps a lot.
There needs to be something like, "I tried, but found it to be overwhelmed with low-quality, low-bid experts-at-everything offshore body shops and astroturf/sockpuppet-portfolio-inflated accounts, which disgusted me and so gave up on it." Ie, horrible signal-to-noise ratio. A cattle call.
I agree there is a load of low quality developers out there on oDesk, but I've found that for the contracts good developers want, the buyers know not to hire the low quality developers. I've found plenty of high paying contracts on oDesk, you just need to know how to approach them.
FWIW, this is apparent in the data as well---employers are price sensitive, but not nearly as price sensitive as people might think. They good employers aren't stupid and realize you get what you pay for. In fact, the number of one predictor (by far) of who gets hired on oDesk is not experience, hours, tests, county etc.---it's how customized their cover letter is (I took the Levenshtein distance between an applicant's cover letter & their previous cover letter to create a "customization score" that I used as a predictor in a model of hiring).
At least for development work, where each job is only receiving ~10-15 applicants, a well written cover letter that talks intelligently about the employer's problem can cut through the noise of low-quality applicants & lead to a hire.
Just wanted to plug in my start-up. For basic data mining tasks over the web you can use application built over top of the platform https://excavator.mobileworks.com and you wouldn't have to hire anyone. More apps coming soon. Email support and we'll be happy to give discounts to fellow HNers.
Seeing the positive response to sites like http://tinyproj.com and http://weekendhacker.net, it seems like highly competent freelancers would benefit from a middle ground between the somewhat impersonal market of eLance/oDesk and the inefficient method of seeking work via personal networks. Something like a "hackers guild" of people who have been vetted for quality work. Projects could be semi-automatically assigned to single or multiple individuals based on the importance of the project as measured by the dollar value attached to it and the skill of the individuals as measured by their historical feedback ratings.
Maybe this is just my naive impression as an undergrad approaching graduation who would prefer freelancing to a 9-5 or even a startup but hasn't come to terms yet with the fact that inefficiency in the matching process and initially low rates are unavoidable. If not and you'd be interested in building or using such a system, I'd like to talk to you. It also relates to emerging things like the "reputation graph" and transitive delegation that can enable these sorts of systems, and are going to be huge (see http://blog.hyperarchy.com, http://liquidfeedback.org, http://rglabsinc.com).