Is matching technology the silver bullet that employers (and some vendors) want it to be?

Merriam-Webster defines a “silver bullet” as something that acts as a magical weapon, especially if it instantly solves a long-standing problem. Sounds to me like the promises that a lot of vendors make to potential customers, including promises made by some HR tech vendors to employers. 

A recent episode of The Chad & Cheese Podcast (note the capitalized The, as in The Ohio State University, but I digress) caused me to ruminate about this subject, and if there’s one thing about rumination that I don’t like, is that it often ends up as vomit. The guest was the bright and likeable Claire McTaggart, chief executive officer of SquarePeg. During the episode, Claire described her company’s mission, product, customers, pricing, and value proposition and then the hosts, Joel Cheesman and Chad Sowash, passed judgment. In short, they liked her but not her business model. There were a number of aspects of her business to like but also some deal killers, chief among them the reliance on matching technology. Joel listed examples of sites which had claimed to have superb matching technology but essentially no longer exist including JobFox, Jobster, and Climber. What the hosts did not take the time to dive into, nor was that episode the appropriate time to do so, was why none of these sites have been able to make matching technology work and why that may be an impossible task, at least for certain roles.

In theory, an employer should be able to upload to a job board a job posting and a candidate should be able to upload to that same job board a resume and the job board should be able to magically match those two together. But that doesn’t happen. Why? Well, let me list just some of the problems:

• Job postings are forward and resumes are backward looking documents. In a job posting ad, an employer mostly describes what they want the employee to do in the future. In a resume, a candidate mostly describes what they did in the past. If the candidate wants to continue to do in the future what they’ve done in the past, then I could see matching technology working, but if the candidate wants to continue to do in the future what they did in the past, then why aren’t they still doing it? Something didn’t work out, right? So, presumably, the candidate wants something to change with their next position. That change could be subtle or major, but there will be a change.
• Matching technology requires a significant investment of time by both employer and candidate. I can absolutely see scenarios where the recruiter is ready, willing, and able to invest 20, 30, or even 40 minutes to answer question after question about a role so that the matching software really understands what the job entails and what qualifications are required for it. And I can also see scenarios where a candidate is ready, willing, and able to invest 20, 30, or even 40 minutes to answer question after question about the role their qualifications and role they desire so that the matching software really understands the candidate. But how often will both occur? You need both to for the software to work. In a job market where candidates currently have the power, they have little incentive to spend that kind of time without any guarantee of good results. When the next recession comes, and it will come at some point, candidates will have that incentive, but employers no longer will. There are rare occasions when the labor market is in equilibrium, but far more times when it is not.
• Both parties need to be well informed and articulate. Employers gripe all the time about poorly written resumes. Candidates gripe all the time about poorly written job postings. The reality is that they’re both right. Many (most?) resumes are poorly written. Many (most?) job postings are poorly written. This reminds me of an expression that my grandfather used to say. “When you put one pig together with another pig, you don’t get a swan. You get another pig.” Okay, so I’ve been blogging for two decades and have tried hard to work that expression in time after time and I finally found an opportunity. So cut me some slack, okay? But, seriously, the matching systems assume that the information entered by both parties is accurate…but it likely won’t be. Recruiters often don’t understand the needs of the hiring manager and hiring managers often don’t know how to describe the candidate they want. “I’ll know it when I see it,” is often heard when hiring managers meet with recruiters. If the hiring manager can’t articulate to the recruiter what the candidate needs to succeed, how can the recruiter accurately describe the desired candidate to the matching software? And, let’s be honest, most candidates for most jobs don’t give a crap about the employer’s brand or products or opportunities for growth. They want a paycheck larger than the paycheck they’re getting now and they want to shorten their commute and hopefully not work for a boss who is a total asshole. So how do you describe that to matching software?
• Privacy is a bitch but we all want it and it is coming. It seems that a rapidly increasing percentage of the sites that purport to match candidates with job openings are doing so without permission from both sides. The job board market is double-sided, meaning that job boards serve two customer groups: employers and candidates. Typically, only the employers pay a fee but both groups are customers. The start-ups in our industry that rely upon artificial intelligence or machine learning understand that their technology only works with massive amounts of data, and they’ve deemed it too expensive, too slow, or both to convince enough job seekers to register and provide enough information during that process for the tech to work. The answer has been to scrape LinkedIn or otherwise acquire data without permission from the candidate. LinkedIn has been aggressively suing these companies. It is too soon to say who will win that battle, but few organizations would relish a fight with LinkedIn even before its acquisition by the much larger Microsoft and certainly wouldn’t relish a fight now. To compound this problem, the European Union’s General Data Protection Regulation (GDPR) and California’s somewhat similar Consumer Privacy Act (CCPA) shift control of data back to the user, meaning that job boards that rely upon data gathered without permission are likely to face repercussions including fines from governmental regulators. To me, this points toward a future where matching becomes more difficult not because of a technology problem, but instead because of a data problem.
• It just doesn’t scale. This may not matter to an employer who hires one person here and one person there, but most of College Recruiter’s employer customers hire at scale, meaning dozens, hundreds, and sometimes even thousands of students and recent graduates for part-time, seasonal, internship, and entry-level jobs. If you’re any of the thousands of employers who hire 10 or more people per role and you have hundreds to thousands of roles advertised at any given time, it simply takes too much staff time to manually walk through an online questionnaire about each and every role. Automation and process become key. Rather than filtering out people before you even connect with them, it becomes more efficient and effective to engage with them, assess their skills, and either continue to move them through or out of your pipeline.

If I were an employer and had a role that required very, very unusual skills, then I might be tempted to use matching technology. But I wouldn’t rely upon it. I would look at that role, realize how critical it was to the success of my organization, and call in a good sourcer and/or third party recruiter (a/k/a headhunter) for it is those people who are trained and capable of talking to the hiring manager and ferreting out details that don’t make it into a job description and then doing the same with the candidate. And, for the other 95 percent of jobs that I needed filled, I would rely upon a job board well targeted to my niche (such as College Recruiter for part-time, seasonal, internship, and entry-level jobs) AND the best matching technology ever invented: the muscle that has always existed between the ears of the candidates and recruiters.