We originally wanted to do this for web developers, but we didn't find it to be very useful without the visual component.
With web design, you can very quickly scroll through a lot of portfolios and find a style you like. That wouldn't work at all with web developers and backend code.
We're going to be launching the index page shortly and it'll become even clearer why this form only works for visual works.
I'm not sure you need the visual component for developers - the hard part is getting a sense for what things cost, how long they take, and how people work.
Turns out those things are really hard to get a feeling for in isolation, but in comparisons, people are really good at it. I know I want an iPhone app; I come across http://www.flingmedia.com/, and the contact us form gives me his rates. Then you try to see if Brian's rates are valid; the pricing itself gives a marker of credibility. If Brian's rates are in line with what I see elsewhere, and his portfolio seems directionally suggestive enough that he can build what he says he can build, at the very least, I'm going to contact him. The conversation with Brian will tell me, as a client, if he'd be good to work with. Sure, it takes more work - you can't make judgments based so heavily on portfolios - but I really think Haystack would be really helpful to both clients and developers. (So, please, add developers!)
Semi-related, I'll bet you a dollar that people will be looking for local designers before they look at portfolios (rather than starting with portfolios) because you give them that option. For developers, I bet they'd do the same. You'll be able to track that relatively easily.
You can be separated by distance and still easily use paper to quickly iterate through ideas. 37signals themselves do this - they're geographically distributed yet use Sharpies to think through rough interface ideas:
All of my freelance clients in the last few years have been startups in the Bay Area (where I live) so I understand all of those reasons and the benefits of proximity. Hell, I wouldn't have moved up here if I hadn't read Hackers And Painters in 2005 - I wanted to be where the action was. But I've been slowly doing less and less on-site work and have gotten to the point where I don't do much more than weekly meetings and not even that with some clients, so I can't help but realize that I could've technically been anywhere in the world for all intents and purposes. And these aren't low-paying entry level portfolio builders either, it's real paying work with the same challenges as always.
So again, are the reasons for working with a local designer more important than overall skill and portfolio?
It's not only about caring more, but also about which filters are offered.
Since "near this city" is filter 1 and price range is nr. 2, people are going to pick their city to shorten their options from several hundreds to a couple dozen to make it easier to choose.
Not too big of a deal right now, but once there are 10x as many designers on there, this is going to be one of those straws clients will grasp on to, even when geography really isn't that relevant to them.
Yeah, you're right, but you can probably also agree that honing in a certain market segment (graphics designers) is a better idea / opportunity than to go into competition with other typical job-related bulletin boards...
"... matching designer’s work so they can hone in visually: ..."
Interesting to see the contrast b/w 'visual' search to search by 'algorythm'. Less work involved, product gets out quicker and works. Will the machine, the algorythm make selection better (easier for users). I'm thinking of Demand in particular ~ http://stephenlaughlin.posterous.com/demand-media-uses-algor...
With web design, you can very quickly scroll through a lot of portfolios and find a style you like. That wouldn't work at all with web developers and backend code.
We're going to be launching the index page shortly and it'll become even clearer why this form only works for visual works.