It was just to rubber-stamp the decision to recruit Joanne (not her real name) for a UK HR director role – a relatively informal talk over Skype with the global HR director to seal the deal. The position was in the bag. Except…
“That ineffective Skype interview cost me a job I could have done in my sleep,” she says. “It had a huge effect on a role I had secured in principle.”
Anyone who has used real-time internet video will sympathise. Four minutes into the call the connection drops. The mad panic to try to connect again. Trying to make light of it and pick up from where you were. The time lag doesn’t help, of course. Neither does the language barrier when your signal is going in and out.
What if this had been a traditional ‘in-room’ interview? Would Joanne have got the job? We will never know. But according to group HR director at CCS Care Services Lucy Dodd, this type of video platform was probably not the right tool at this stage in the process. “I would never do an interview over Skype for a senior role and certainly not for a final interview. Face-to-face is important at this level,” she says. “I think most people tend to see Skype as more informal, for catching up with family members, for example.”
And yet at the Institute of Student Employers, 49% of members now use video interviewing as part of the recruitment process. Such interviewing may take the form of a live face-to-face, as in Joanne’s experience, or a recorded, on-demand interview. So, does this mean the death of the traditional in-room interview? And, potentially, more disastrous scenarios such as the above?
The problem with the traditional, face-to-face approach to interviewing is that it is flawed, particularly when it comes to predicting future job performance. As research by Frank L Schmidt and John E Hunter found in 1998, the interview is only ninth out of 19 different selection methods that can be used in hiring when it comes to validity for predicting future performance – well below general mental ability and work sample tests (that is, tests related to the tasks the job candidate will have to perform).
However, validity very much depends on whether the interview is structured or unstructured. In Schmidt and Hunter’s meta-analysis, one of the most powerful predictors of job performance was general mental ability combined with a structured interview – one where the questions are asked in a set, standardised order and where the interviewer does not deviate from this schedule. A subsequent paper by Schmidt, Oh and Shaffer in 2016 also found the top three predictors of job success to be work samples (29%), cognitive tests (26%) and the structured interview (26%). All three test how someone thinks and problem solves – good indicators of likely effectiveness in the job.
Technology is well suited to delivering a structured interview, and the latest on-demand video platforms do just that. Such technology has flipped the hiring funnel on its head, with organisations using video interviewing as a screening tool at the beginning of the process. Candidates record themselves answering pre-determined questions in their own time and at their own convenience on any device, saving travelling costs and providing a much better experience. (Recruitment provider HireVue claims a net promoter score of 70 for its video candidate experience.)
Of course there is no reason a human cannot conduct a structured interview. Yet, says behavioural economist and professor at the Harvard Kennedy School, Iris Bohnet, unstructured interviews – not structured – consistently receive the highest ratings for perceived effectiveness from hiring managers.
‘The unwillingness to give up a much-loved evaluation approach seems to be driven by two factors: managers are overconfident about their own expertise and experience, and they dislike deferring to more structured approaches that might outsource human judgement to a machine,’ she wrote in a Harvard Business Review article earlier this year.
Not only are unstructured interviews poor predictors of performance, but they are also fraught with bias and irrelevance. However sensitive to the issue of unconscious bias a recruiter may be, as professor at the Yale School of Management Zoe Chance says: “Bias is an inherent part of us. Just knowing about it doesn’t change it. We can’t get the bias out of the person so we have to get it out of the system as much as we possibly can.”
Bohnet agrees: “We should stop wasting resources trying to de-bias mindsets and instead start to de-bias our hiring procedures.”
This is where technology-based interviews can make a difference. When Philip Wilson, chief psychologist and chief assessor at Civil Service Fast Stream, looked at metrics from his Oleeo video interviewing platform, he found something he was not expecting. “The video interview was very successful in achieving positive diversity outcomes across categories. In some cases, the minority group outperformed the majority group.”
He puts this down to a more structured approach. “I think the strengths-style questions supported this diversity because they focus less on historical evidence, while the video interview platform brings in a fairness by making the assessment process consistent for all. Above all, perhaps, the standard questions across candidates remove bias.”
While the use of data-based technology has come under fire with the recent news that Amazon’s tool using artificial intelligence (AI) to sort through job applications resulted in sexist outcomes, HireVue’s chief technology officer Loren Larsen says the number of diversity hires improves using AI-based selection, which removes biases inherent in the human interview.
“We built a machine-learning model to see if we could predict attractiveness ratings and whether this affected hiring. We found that, if someone was rated a seven, they were twice as likely to get hired as someone who was a three. But when AI was used, the model didn’t pick up attractiveness as an issue at all as it was not relevant to the job,” he says.
Data can be crunched to recommend the best questions to ask for a certain role based on the competencies and skills required, and then algorithms capture information from the video interview. This can include information such as words used, facial expressions, how people say particular things, if their language is definite or moderate, if they speak in the past tense all the time, and where they put emphasis. Depending on the job role, these may be important or not, but data such as this can be used to build an algorithm to predict success in the role.
Larsen says HireVue has captured data from 6.3 million faces from 87 countries. It is now going a step further with the integration of gamification into its on-demand videos. Candidates receive a video invitation, answer questions on the video and then play a couple of games. Cognitive assessment is combined with deep behavioural insights.
According to Jennifer Carpenter, VP global talent acquisition at Delta Airlines, on-demand video interviewing has not only reduced costs, saved time and produced a better pipeline of talent, but has also helped the US-based airline company to deliver a better candidate experience.
She explains: “We have one million applications a year for fewer than 10,000 jobs, so 99% of people are disappointed, and they are also our customers. We’ve now added the ability to re-record to our videos. So what if people are over-prepared? We have asked candidates to make our lives easier by talking into a phone or laptop and telling us about themselves. So, we want them to walk away from the experience feeling comfortable and confident they have done the best they can. About 60% of candidates are opting to re-record and we have seen our NPS [net promotor score] increase by seven points on the back of this.”
It is unlikely technology will remove the human face-to-face interview altogether. But what it will do is give an objective way to assess candidates and deliver the best people to the recruiters and hiring managers, so the human interaction that follows will be more successful. It can also help deliver a better candidate experience. So interviewing is here to stay. The irony is that cutting-edge AI technology may just help make the whole process more human-orientated.
This piece featured in our What's on the cards for hiring? ebook in partnership with LinkedIn. Read the full supplement, including extra box-outs, here