Why are so Many Applicants being Rejected in Times of Candidate Shortage?

Around 80% of Australian employers and recruitment firms use an Applicant Tracking System (ATS) and there are over 200 to choose from. But is this rush to ‘improve’ the recruitment process actually creating more problems than solutions? In times of candidate shortages, it may be that humans are smarter than robots after all.

Bernadette Eichner

Sep 20, 2021

Share:

There is a lot of talk about the shortage of skilled candidates in Australia right now. It doesn’t matter who you talk to – recruiters or internal hiring managers – the lament is the same – “we just cannot find enough good candidates right now”.

Yet, we often hear from very skilled and qualified people who are applying for multiple jobs and not even getting their applications acknowledged as having been received, let alone an interview opportunity. And TV news items are regularly interviewing people who have been stood down, laid off or terminated due to the COVID restrictions and saying how hard it is to get a job at the moment. These people are well qualified and experienced people who are applying for multiple jobs without success.

So what is going on?

Could it be that applications aren’t even being received by the recruiter or hiring manager?

A recent study by Harvard University suggests that more than 10 million applications never see the light of day due to being rejected or blocked before they get to the hiring manager’s Inbox. While this was an American study, there are similar findings in the UK and Australia.

Increased use of AI technology in the recruitment and hiring process has made it easier for people to apply for jobs but it has also made it easier for companies to reject them, without even sighting their resume.

One of the most popular software systems around, collectively known as ‘Applicant Tracking Systems’ is one that sorts through applications and rejects or accepts them based on a set of ‘rules’ the company has put into the system. These ‘rules’ can be anything from rejecting any candidate that doesn’t have a particular degree, or doesn’t use any key words from the job description, to anyone who has a gap of more than 3 months in their resume.

No matter that the applicant may have a raft of transferable and perhaps even desirable skills and experience to offer.

Around 80% of Australian employers and recruitment firms use an Applicant Tracking System (ATS) and there are over 200 to choose from.

And to add insult to injury for job applicants, some systems also reject resumes that may contain different formatting or use underlining, italics or bold text to emphasise something, so the application doesn’t even make it into the system.

The idea is that these systems can speed up the recruitment process by only letting ‘the most suitable’ applications’ through for review. The ‘most suitable’ being the ones that comply with all the ‘rules’ and use the right words.

But are we shooting ourselves in the foot?

Is AI actually rejecting suitable candidates?

Earlier this year a client asked for our help with a very senior specialised role, but given the candidate shortage challenge wanted to run their own advertising campaign in parallel to maximise what we in the business call ‘candidate reach’.

Of course that was fine with us (we’re good like that) and we agreed that they would provide us with a daily list of applicants they were considering so we weren’t spending precious interview time with the same candidates. They were confidently using their ATS to screen applications and reject those that were unsuitable.

In our search activities, we found a candidate who was, in our opinion, far and away the most suitable for the job. In fact, she was spot on, having done the same role for a similar type and size company in the past.

We checked with our client that they didn’t have her application on their desk and got a clear ‘No’, so we reached out to her. She mentioned she had applied directly to them but hadn’t heard anything back. Again, we checked with our client and again, got told they didn’t have her resume.

So, we presented her. She was superior to any candidate sourced to that point and was interviewed and offered the job within days.

It was later discovered that she had applied directly to our client but been rejected by their ATS so her resume never found its way into the Hiring Manager’s inbox. Why? Because she hadn’t worked at one of the five companies the ATS recognised as preferred experience.

AI algorithms are typically trained on past data, so if you don’t have diversity or flexibility in your data set, how can you possibly know how individuals from underrepresented groups may have performed in the past or how they would perform in the future?

For example, many systems reject applications for reasons that could be explained if the candidate was given the chance to. A common reason for rejecting an application is gaps in employment. But it may be that someone was taking time out of work to have a baby, or care for a terminally ill spouse or child, or deal with a temporary health challenge of their own. None of these reasons make them unsuitable for the job they’ve applied for. Or it may be that they moved states or countries and needed the time to settle into their new home and surrounds before starting their job search. Again, this doesn’t mean they aren’t competent.

Examples we’ve seen here in Australia include applications for multiple customer service roles in a large online retail company where the software only accepted candidates with experience in ‘e-commerce’, when all they needed were people who were able to use simple CRM and telephony tools. As a result, only two applications got through the system when there were scores more who met the criteria for the job.

On the flip side, we’ve seen companies be inundated with completely unsuitable applications because the system accepted anyone with the keyword ‘GPS’ in their resume, when the role was for a specialised technical sales person in a company that uses GPS technology in their own bespoke systems.

Should companies stick with, or scrap AI in the recruitment process?

One of the biggest software companies in the world, Amazon, scrapped the use of AI systems in the recruitment process way back in 2018 when it became obvious their system was rejecting women for the types of jobs that had usually been held by men. As it turns out, that was just the tip of the iceberg and they realised they were losing access to outstanding people who could make a valuable contribution to the company’s growth and success because the algorithms couldn’t apply any level of ‘consideration’.

Many others have done the same.

We believe that AI has its place in the recruitment process and can certainly make the process faster and smoother. But we stopped relying on our ATS years ago to tell us who the best candidates were. Yes, sure we do a quick search and sort but then we throw out any limiting criteria and look at the list again.

In times of candidate shortages, it may be that humans are smarter than robots.