Rage against the machine: Automation in recruitment is failing us all
Pursuit of volume alongside outdated processes is growing the tech skills gap
In the second in our series on the tech skills gap we examine the gap from the perspective of jobseekers and academia and try to answer the following question: If there is such a huge tech skills gap, why is it so hard to get a decently paid entry level job?
Last week Computing examined the tech skills gap from the point of view of employers. Whilst we found some agreement that graduates lack the ‘soft skills’ necessary for the workplace, there was less agreement on who is responsible for fixing the problem. Should universities ensure that graduates are ready to slot, fully formed into entry level roles? Or should employers take responsibility?
Nicky Danino is Head of the School of Computer Science at Leeds Trinity University and Chair of the Council of Professors and Heads of Computing. Danino is also on the British Computer Society (BCS) accreditation committee. She acknowledges the soft skills problem.
“The consensus from employers that I've spoken to seems to be that they don't feel that graduates are ready to hit the ground running. Coming from a higher education point of view my answer is ‘of course, they're not.’ I think we need to manage the expectations of employers of what a graduate is capable of. Degree courses are not work training courses.”

The main issue is that academic institutions and degree courses must focus limited resources and time into core, academic skills.
Mark (not his real name) is in his final year of an integrated master’s degree in data science at a Russell Group University. He describes how the financial pressure on universities affects outcomes.
“Universities have a far greater profit incentive than they used to because they require a certain number of students to turn a profit that can, in turn drive their research. Even since I joined the university, the drive has been geared towards getting more students and it tends to be international and postgraduate students that pay the biggest fees. I think universities are stretched and aren't really providing the services that they used to in terms of equipping people with skills.”
Whilst acknowledging the difficult financial situation in universities, Danino says that contrary to the line that universities focus solely on theory, what she prefers to call transferable skills are considered in course design.
“If you want BCS accreditation there is a focus on things like social, ethical and professional issues. You cannot get a course accredited unless it includes group work, and they look at whether students are being taught critical skills. They are looking to see whether we can create that well rounded individual.”
Valued experience
According to Danino, an idea raised in academic circles is that of a post-graduate training programme to ready graduates for the workplace. That sounds great, but who pays? On this, Danino is clear.
“Employers need pay for it, because if employers want somebody with more experience, they’ll end up paying more in wages anyway. So could they redirect that money and use it to train somebody straight out of university.”
What employers consider fair graduate pay is a sore point for graduates. UK median graduate pay peaked in 2008 and has stagnated ever since.
The website talent.com gives a median average computer science graduate salary of £28k per annum. For comparison, a minimum wage role would pay an annual income of approximately £22k per annum.
At this point, it looks obvious that employers are going to struggle to fill the limited entry level roles they have if they demand years of experience and the transferable skills that this experience helps to form, whilst offering relatively low pay.
Despite the fundamental unfairness of this position, students such as Mark are finding ways to gain experience as they study.
“I'm always looking for gigs and ways to expand my professional skill set,” he says. “After my first year of uni, I did a four-week unpaid internship. After my second year, I did a paid summer placement and the same in the summer after my third year.”
In addition to the holiday placements, Mark has had a continuous part-time paid job for 10 – 15 hours a week doing coding and data work for a start-up since his first year.
Neha Yasin is a final year Physics student at Imperial College London and an enthusiastic STEM outreach advocate. Yasin is a Stemettes Futures Youth Board member and Stemettes Social Media Advisor. Regarding work experience she says:
“Everyone I know doing a degree spends their summers doing internships and work placements instead of actually relaxing or doing other jobs.”
Less scrupulous individuals and companies are making hay out of students keen to acquire experience. Nicky Danino says:
“I have had people call me and say ‘I've got a great idea for an app and it's going to make lots of money. Have you got a student who needs work experience to help me build it?’ My first question is always how much they plan to pay.”
The answer is sometimes nothing, and the reason given for that is students needing experience. It looks exploitative.
“I’m not a fan of unpaid work for computing students,” Danino says. “If a student is adding value to a business, they should be getting paid. You should be paid according to your experience and your expertise, but you should not work for free.”
Mark will finish his course this summer and is job hunting in earnest. His impressive work experience – both paid and unpaid - does not seem to have made the difference he hoped for.
“I’ve been job hunting probably for about 18 months. In terms of open, formally advertised roles, I’ve applied to about 34 and sent another 15 to 20 cold emails to companies I'm interested in. I’ve also applied to about 10 recruitment agencies and only heard back from one. They said they would keep my stuff on file.
“I was rejected from four of the 34, I’m awaiting decisions from another three, and I was invited to perform some tests for another three. A few of these might still be in play, but most of these applications were made between September and December and I haven’t heard anything since the start of the year. Some of these have been radio silence for three, four or five months.”

Computer says no
Mark explains the test process. Recruiters begin with verbal and non-verbal reasoning tests then move to personality testing. If you get past that you will be invited to demonstrate coding or design skills in more role specific testing. Beyond that you move to the assessment centre and/or video interviews. I ask Mark at which point in the proceedings you get in a room with other people.
“I’ve never had an in-person interview,” he says. “I know some people that have ended up at in-person interview stage but that’s right at the end of the process, maybe about the fourth or fifth stage.”
Nicky Danino doesn’t think that the recruitment process that Mark describes is fit for purpose. She says:
“To recruit technical people, we tend to follow this traditional method of application, technical tests and interview. Lots of people who work in the profession don't really fit well into these traditional processes. You might be missing somebody who would be a great fit for your company, who would bring a lot of value to your business, because the method of recruitment doesn't work for them.”
Mark puts it more strongly.
“My overwhelming feeling is that this entire exercise is deeply inefficient and flawed. I’m part of the data science society at my university so I speak a bit to employers as well as students. All the employers say, ‘we send out a job, and we get 150 applications, and 95% are no good. We spend so much time just churning through CVs that aren't suitable for the role.’ Then anytime I speak to someone applying for jobs they say, ‘I've applied for 150 jobs and 95% never called me back.’ These are people going to a Russell Group university, doing master's degrees in a quantitative subject and they can’t get call backs.”
Mark blames the recruitment sector for this inefficiency.
“Recruiters, the recruiting software, the companies that offer these AI screening packages built around algorithms and video interviews mean the industry has moved away from quality to volume. They moved away from having a network, understanding clients and getting to know people. It’s now a highly automated process where profit comes from volume. They want as many job applicants as possible, and as many open jobs as possible at any given time to maximise use of their products.”
Recruitment service and products are supposed to find the best people for companies to enable them to grow their businesses. Right now, the processes and products being utilised are failing to deliver that outcome. Recruiters have replaced accrued knowledge of their clients and candidates with automation. This has led to the tech skills gap getting bigger, and an army of under employed STEM graduates working in jobs that would once have been filled by non-graduate school leavers. Neha Yasin shared her sister’s story.
“My sister graduated in neuroscience a couple of years back, and she could not find a job in the field. She ended up working in events for a data company. It's a very common issue.”