Why are niche job boards still emerging in a mass customization era?

This week, Joel Spolsky (whose book Joel on Software is required reading for web entrepreneurs and product managers) launched a job board. It got me thinking about the fact that TechCrunch has a job board, paidContent has a long-running and successful pay-per-post job blog, and, among other sites I consume regularly, Fred Wilson has, instead, an Indeed.com badge configured to list start-up jobs. So my question is, Why do we need niche job blogs when we have highly searchable, filterable, supposedly better, horizontal sites like monster.com, and aggregation/meta-search models with all the web2.0 technology you could wish for like Indeed.com?

Answer: because they work better! Among the 3 niche sites mentioned above, I can find, without hitting a single “next page” button or filter, listings for Director of Philanthropic Relations at Squidoo (wow!), a Product Manager at Google, design guru at Internet Brands (CarsDirect, etc.), and various other VP- and Director-level positions at startups and large internet players all over.

On the other hand, the most interesting job on Fred Wilson’s badge - which is basically a persistent search of jobs with the word “start-up” in them and do not have the word “engineering” in them, a reasonable-looking proxy for non-technical start-up jobs - promises a position as a product manager at a Sequoia funded company, but links to a dead craigslist homepage. Clicking through the badge to see the full list of jobs, one finds 24,250 results for mortgage loan originators and “start-up opportunities” requiring skills like:

candidates will have 3 years or more of progressive experience in operations and customer service, a working knowledge of DOT safety requirements and superior communication and organization skills. Candidates must also meet the following requirements: Have a CDL Class B w/Tanker and Haz-mat endorsements; ability to install propane tanks; excellent customer service, interpersonal and communication skills; and must be qualified for Propane Tank Installations and Bulk Deliveries.

Sorry guys, the only propane installation I’ll be doing any time soon will be for the BBQ I’m having this weekend! And filtering 24,250 jobs has all the hallmarks of a low effort-to-reward exercise when the sliders are inefficient, non-normalized job titles.

On the employer side of the equation, Joel has a good analysis of why it’s hard to find the best people on job sites (nutshell: because they’re rarely seeking jobs). So you need to find them in the places they congregate for other reasons. . . such as reading useful blogs like Joel’s.

This is interesting because it has everything to do with the social dynamics that are behind social media - there’s a delicate balance they must maintain between selectivity (filtering noise) and value (receiving signal), and that balance differs across social groups and functions. LinkedIn is of value to me in business if most of the people in my network are similarly thoughtful to me about who they accept as contacts (it creates a reputational system), but it is not useful to me in finding stuff to buy from trusted parties like ebay is (because there is no marketplace associated with it). Myspace is of low value in maintaining business relationships, but great for bands with a following. paidContent shows that job boards can scale effectively to large audiences, but only because there’s a pricing engine (and possibly an editorial one) maintaining a filter on the thousands of junk listings that would turn up there if they were free. On the other hand, craigslist is a much better place to find apartments for rent, because you need all the signal you can get there and can efficiently filter lots of noise via parameters like price, bedrooms, and location.

Blogs continue to turn out to be great proxies for niche social networks - if a place that I find valuable for information becomes a meat market for jobs, that’s great for me as both a potential employer and employee. On the other hand, the more spaces like this that emerge, the more I have to spread my job listings and consumption across multiple sites. So how do we aggregate this more efficiently?

Hello friends: how about a feed network for job blog sites? How about a Hype Machine for job listings? How about a utility from technorati or Newsgator or MyBlogLog that can identify sites I’m reading or getting readership from or have audiences like me and then surface me feeds of certain kinds of links (podcasts, job listings, product links, Flickr pointers) from my OPML file or community?

What else?