There’s been so much in the media lately about our privacy and data as individuals. Documentaries, films and social media gossip tells us that data is today’s gold, and as such, dubious decisions and ruthless campaigns have been exploited or carried out behind our backs. Data can be lots of things – even a weapon in certain hands.
Data is valuable to recruiters, too, as it can help filter out candidates not so suited to a role, saving time and money during the hiring process. The ‘computer says no’ feeling is indeed disheartening for candidates when job-hunting, but surely it’s better to find out earlier rather than later that you don’t meet written (and, perhaps, unwritten) guidelines and specifics about the job in question, and are therefore free to move on to the next challenge?
According to research, for top-level, sought-after roles, data sifts out 95% of the hundreds of applicants who invariably apply. Rather than using simply fixed/rooted data (such as date of birth) that will never change, technology used for data harvesting within recruitment is increasingly intelligent; it improves its processes and adjusts to ever-changing key indicators as time moves on.
There are some benefits to a computer making the shortlisting decisions. Whilst it may sting that you’re not able to plead your case to a bunch of bytes, AI won’t judge you unconsciously (or will it? Read on…). It certainly won’t assume things from your appearance or take an instant dislike to you.
That’s not to say AI is perfect. It’s borne of data, and not all data is equal. Human intervention will undoubtedly be a huge part of any algorithm or precepts such technology uses. The program itself will have been written by a human, which doesn’t automatically eliminate the existence of any bias. Also, any learning or adaptation by AI will likely be slow, whereas a human could immediately alter their questioning or approach in situations. Such a case was taken to court where the applicant, who had Asperger’s, won her case. The recruitment web design algorithm she’d been using to apply for a role rejected her short, factual answers as it had no understanding of her abilities. Unwittingly, it had unlawfully discriminated against the candidate.
GDPR rules are quite stringent on the use of data. Some relate to decisions made solely on automation without any human intervention; does this mean, therefore, that any automated sifting, profiling, filtering or shortlisting comes under this legislation? Wouldn’t that render the whole process useless?
Experts recommend that the collection of data be an ongoing process. They say that an algorithm learns more about a role when it’s in action, in order to better understand its unique aspects and features and subsequently appoint a suitable candidate. The better the data fed into the program, the more likely the right person will be found for the role when hiring is needed.
Such programs also need to understand the company’s values, indicators of success and future plans. As the talent pool continues to shrink and companies compete for the best candidates, your program needs to know how to identify future leaders and innovators, not just those only concerned with getting the job done. It’s a big ask from a machine, but the larger companies seem to be managing it.
Your website is the first port of call for any recruiter. Setting the required tone and detailing the relevant specialisms and industry the recruiting agency represents is the very first – if unconscious – filter before any programs or algorithms even get going. The first hurdle. The most important information on a virtual plate.
Does your website accurately present your company?
