Hacker News new | ask | show | jobs
by ismarc 5940 days ago
Here's the problem with the current "employment opportunity discovery" mechanism. You need a large number of qualified applicants to review job postings and to reply to those, generating a "success" for a particular site for a particular company. You need a large number of accurate, quality job postings to attract applicants.

I've thought about this space for quite a while, fragmenting job sites all charge money to the company just to post an ad. From that, you expect a user response. I think the ideal method would be to have an indexed, structured database of potential applicants, people who are interested in potential job offers. Company's then input their requirements and a list of potential applicants is spit out, no identifying information, just the raw skills/experience (10 years with C++, 15 with C, 2 with C#, etc.). The company then pays $x to solicit that potential applicant into applying for the job. This preserve's the anonyminity the current system has (no personal identifiable information revealed until application takes place) and allows companies to spend significantly less time attempting to discover potential applicants. It's not without it's massive drawbacks (you need the index of applicants, get companies to use it, get applicants to keep their information up to date, gives the applicant a feeling of not trying to find a job, HR does the hiring so they may not pay for the grey area applicants when normally they'd at least make it to the hiring manager) but you can overcome some of these by providing the traditional service as well, company's pay a "nominal" fee to list a position and applicants search/apply.

From what I've discovered, the best jobs are the ones you can't find. They're posted on the companies job listings and sit there, or the current job search systems have crappy indexes and include more cruft than quality (try a search for 'c linux' on one for a good example). Hmm...maybe there's room in this sector if one could tap into a large number of people who are interested in anonymously being considered for other potential job opportunities.

2 comments

Your last paragraph is on point.

I disagree though with your idea about a database of applicants thats searchable by skill.

I think your system would fail without a means to authenticate and validate people's credentials, plus without a validation layer it'll only encourage people to exaggerate or lie about their skills.

The other problem from the perspective of an applicant is submitting resumes and cover letters to online postings often feels like all work for no gain. Most of these postings feel like black holes. You submit something and (typically) never hear back.

Maybe an improvement to your idea is some carrot/stick mechanism, that at the very least, incentives both parties to stick to good behavior (timely submissions of RELEVANT resume / cover letters and timely responses from the companies that listed the original posting). This would solve for some of the frustration these sites illicit.

Yeah, validation layer was one of the many, many issues with the system. Maybe a dating site type system...have an index of jobs and applicants. Free to send messages back and forth, completely anonymouse. Messages are "request for application", "application", "denial of application", "denial of request for application", "application accepted" (not you're hired, but the application was received and we're reviewing it), "request for screening" and "accept/decline screening". You can't send a message if you have one pending a response. Make it 100% free, charging money to increase the number of pending responses you can have and still send a message.

Don't mind the rambling...this is literally typing as I think.

Indexing large amounts of candidate data, including extracting 'skill' information from CVs, is pretty much a solved problem. It's what my employer's products do.

We also have products that do 'Candidate Matching' - which is exactly what you're talking about here, taking structured requirements for a specific job and then seeing who in the database matches. We normally do it against a recruiter's own local DB, and can do some degree of matching against internet job boards, LinkedIn, that sort of thing.

There was some interest in a product that would index all of the job boards for every UK local authority/council for one client, but that fell through. Going through every company's individual job postings is just impossible, and wouldn't help anyway: in my experience the best jobs are the ones that say 'no agencies', and it's only recruiting agencies who are spending on the kind of software we make.