Job Matching is Not a Magic Bullet

The most important thing about a job board is that it introduces candidates to employers, and vice versa. If that function is not working, then you do not have a job board.

The basics of how this process works usually goes something like this:

1. Employer posts job opening
2. Candidates discover the position
3. Candidates apply for the position
4. Employer reviews applications
5. Employer contacts the candidates they’re interested in, followed by interviewing, hiring, etc.

(Variations of this scenario include the employer or a recruiter discovering candidates on the job board, contacting/interviewing them, and making an offer.)

How your job board delivers this service is the big question. Some of the answers the industry is supplying lately sound like:

• “Job-Matching”
• “Algorithm”
• “Tinder for Jobs”
• “Predictive Hiring”

All of these methods put the focus on filtering the candidates who reach out to employers. The filters can include keywords for skills, location, level of education, job title, industry, legal eligibility to work, etc. They might also include preferences and expectations for things like salary and travel time. Some methods might also involve personality assessments and aptitude tests.

And these can be valuable ways to organize applications and learn more about candidates as you evaluate them.

But. But. Job-matching is often paired with the caveat of eliminating search. The premise is usually this: “we’ve created an elaborate system for assessing compatibility between jobs and candidates, and our algorithm will serve up matches for you, so you don’t have to do any work.” (Except for the actual interviewing, negotiating, and hiring.) The justification is to spare both the employer and the candidate time and energy by eliminating the work involved in searching.

Personality tests can help you get to know someone better. Matching is great for introductions you might not otherwise have had.

But not at the expense of searching. Searching or refining your “matches” is what you need to do to accommodate the human error and/or oversight inevitably involved both in the design of algorithms or matching technology, and in using them.

One of the most common complaints about Applicant Tracking Systems is that they filter out quality candidates — because of keyword numbers, format, or poor decisions that went into designing the ATS in the first place. Job-matching falls into the same traps unless they allow job seekers and employers flexibility to search, refine, and research beyond the limits of the matching algorithms.

Let’s recognize matching tools and algorithms for what they are: useful but flawed and imperfect tools that can be helpful or harmful in hiring.

I’m not suggesting that reliance on job-matching will result in a horrific dystopian future where we have no choice but to follow the command of heartless, brainless robots. And nor do I think that there is no place for structure in a messily human process like hiring. But take the claims of perfected job-matching technology with a grain of salt.

Recruiting and hiring is hard work. The role of technology is to remove barriers, reduce friction, and accomplish tasks with greater ease. It’s about making the work for humans easier, not trying to take the human out of it.