On the contrary. Greg's work is creating a new market for expertise. If you want to sell your time in 20 minute intervals, it's quite difficult to do so today.
A counter example:
He mentions graphing data as one option. I'm pretty sure people with excellent GNU plot skills can make a GNU plot script that does what I want in 10-20 mins, and I really can't be arsed to figure it out myself everytime. Seems like the type of expertise that is currently hard to sell and might benefit from this sort of site.
As you can tell from that article, I am all for consulting in short intervals, BUT there are a couple issues with the Pomodoro Hiring Experiment. First, 20 minutes is way to short. I have had hour long sessions with several experts since my original blog post and I truly believe that you can only get value from an expert in under 60 minutes if you already have a prior relationship with them and they already understand the basic problem you are trying to solve. If not, it is going to take 10 - 15 minutes just to get on the same page with them.
Look, talking to an expert is not magic. I don't care how good someone is, they aren't going to solve your problem at the drop of a hat even if it is a simple problem just because of the time it takes for context switching and getting up to speed with your background.
One last thing. There is a big difference between getting expert advice for higher level strategical or design questions vs. solving an actual coding problem. I think if you are just using an expert for strategical/design issues, you can get a lot of value in an hour. If you are trying to solve a coding problem, you run a serious risk of not getting any value at all for your short time period. How many times have you debugged an issue that you thought was easy, but then ended up taking you 6 hours. There is little chance you would even ask an expert in the first place for a simple problem (i.e. it wouldn't be worth it), so there is likely a pretty good chance that if you are looking for an expert to solve a specific technical problem it is going to take more than 20 minutes.
In summary, an hour of expert time is awesome for higher level design and strategy but short periods of time to solve difficult specific technical problems will never work.
To me hiring a professional for 20 minutes might sound enough for some small taske, but I'm not sure how well this goes for more complex tasks (e.g. where the pro has to read and understand your code). I guess you can buy couple of chunks than but I'm not sure I see the benefit. Is it just so that you don't have to pay every starting hour?
From the perspective of a person hiring, I'm sure it's better to hire for exactly as long as you need. The point of 20 minute chunks is to make the idea more appealing to experts (who are busy doing other things). As an expert, you know that if you accept a job, you only need to commit up to 20 minutes, and you'll get paid for a full 20 minutes (even if they're not all used), to justify the 'switching cost'.
Love this idea. There have been plenty of times when I wish I could throw a few questions at an expert in a given field. "Is this the best way to program an XYZ?" "Can you think of a better way to do this?" Is the price is right, throwing a small amount of cash at a problem is often a better solution than (a) doing something wrong, or (b) spending significantly more time figuring out how to do it right.
I love this. Not only will this be be great when you have those "stuck" moments, or you want to learn Ruby but wouldn't know to start by reading _why's Poignant Guide, but I bet you see some incredibly creative solutions because the expert knows he needs to deliver in a strict 20 minute window.
"Man built most nobly when limitations were at their greatest."
— Frank Lloyd Wright
I'd be pleasantly surprised if he (and oDesk, by extension) were to share this data. They're very profitable, and this type of information might offer them a competitive edge. Moreover, it appears as if they're also fervid believers in intellectual property law's "right to exclusion" applied to ridiculous vague quasi-inventions, viz. http://www.google.com/patents/US20060284838. Luckily, this is only an application, and its longer-than-typical prosecution history will hopefully prevent them from trolling too much, thanks to prosecution history estoppel.
if anything, odesk is among the most open companies i've ever seen (note, I don't work there). They publish lots of data in oconomy and in blogs.
I think Greg's idea is really interesting. It takes advantage of a distributed economy and gives experts the chance to make a few extra bucks. One of the things that will be really interesting is how they will curate the experts on the site to make sure the 20 minutes are with someone who can actually add value.
So I'd assume 20 minutes of an expert's time is worth $500, right?
Edit:
Seriously, I'd charge $500 if the client picked the time and $100 if I could pick any period during the day, on short notice.
I'd actually charge more for a client-picked 20 minute period than I would for a one hour period because a 20 minute period would be very unlikely to repeat since 20 minutes isn't really enough to demonstrate great expertise.
