If you are seeking employment, the first place you go to is the internet. The marvel of online searches is that you can maintain 100 per cent anonymity, if you so choose, and can conduct an aggressive campaign in the privacy of your home without revealing your plans even to your loved ones.
There are so many job sites that it is hard to even list them. Monster, Glassdoor, ZipRecruiter, Indeed, Dice, LinkedIn and CareerBuilder are all popular with job seekers, and rightly so. Each site claims millions of hits every month and promises to do the tough matchmaking between employer and future employee.
The first day that you fill out your profile on each site and upload your standard resume, chances are that you’re feeling tremendously positive. Based on your preferences, each site suggests multiple positions and you even apply to a few of them. You shake your head in wonder at the ease with which you upload a standard cover letter and modify it for those hiring managers you really want to impress. Or you could conduct job searches the Amazon “one-click” way.
Except that you realise very soon that these sites are a huge waste of time. Many posted jobs are not all that accurate; they could’ve expired weeks prior, but the sites don’t reflect the latest status.
In Western countries, some job descriptions are so detailed that you get disheartened even thinking about applying for them. These are jobs that are posted by employers simply to meet a government requirement. They already have able foreigners working in them and they would like to retain these employees because they have invested in their growth. But governments force them to advertise for these positions to make them open to the general public under the theory that if no one qualified applies, the foreign workers can be given legal permission to immigrate. And so the job descriptions are crafted with care to closely match the qualifications of the individual already in the job.
Malfunctioning algorithm
In over 20 years of development, job sites haven’t changed much at their core. The algorithms used to match employee job preferences still largely rely on traditional keyword searches; which is why the notifications of jobs that arrive in your e-mail inbox are largely unrelated to what you are looking for.
Technology jobs are especially difficult to find. Someone looking for a job in software testing or quality assurance is repeatedly served up jobs in quality control at a manufacturing plant or pharmaceutical lab (the only keyword that the algorithm found to match was “quality” but that was good enough for the site). When the algorithm does recognise that the person actually works in the technology field, it still promotes tens of unrelated jobs as recommended — those in different areas of the software development life cycle, such as development, design, architecture, Project Management or Scrum Master.
People just starting out have it really bad. A person seeking an entry-level position — defined in the industry as someone with zero to two years experience — regularly receives notifications of senior openings, those with over eight years of experience. A candidate looking for jobs in one geographical area of the country repeatedly gets phone calls from recruiters about positions in cities hundreds, sometimes even thousands, of miles away.
AI and bots
A recent innovation of the job sites is to make the e-mail notifications appear personal, a theme borrowed from Amazon’s Alexa or Apple’s Siri. This means that you don’t get e-mails from ZipRecruiter — but you do so from “Phil” at ZipRecruiter. While a personal e-mail is a nice touch, the recommended jobs are still way off. Each e-mail from Phil ends with a plea: “Is this job a good match? If not, please let us know so we can improve our results.” This is because Phil is an AI bot who has no human qualities at all.
Finding a match is just the first step towards obtaining a job. This is followed by HR screening calls, the actual job interview and if successful, mutual salary negotiations which culminate in the candidate starting to work at the company. Phil can’t help with any of these vital steps.
Artificial intelligence and machine learning technologies, where algorithms self-learn and improve search results over time, are the promise of many industries, including the job-search sector. Large companies, such as Google and Facebook, use these technologies every day in the hundreds of millions of searches that people conduct on their platform.
Good ol’ human nature
But these technologies are not as mature as one may think. Noah Smith, a Bloomberg Opinion columnist and former assistant professor of finance at Stony Brook University, likes to quote economist Robert Solow who says that “you can see the machine learning age everywhere but in the economic statistics.” By this, he means that “there’s no evidence that machines are taking many of our jobs yet.”
This is why job searches for many are still old-fashioned in nature. You have a much better chance of landing a job if you know someone who can recommend you to a hiring manager, than if you know all the tricks at a career website. Corners can be cut if necessary because in the world of employment, trust and attitude often are more important than aptitude. Phil can never learn the basic human concessions implicit in making a deal. In many ways, we’re back to the glory days of the 1990s.
Phil won’t be happy about this. But Phil, as is his nature, just doesn’t care.