Human hiring using interview automation can save you time and money. It also has the potential to help overcome bias in recruiting and delight candidates. Not to mention help recruiters take back their lunch hours and evenings.
So how do you get started?
Here’s the complete guide to human hiring using interview automation.
Interview automation is the process of having tasks which would ordinarily be done by people, done using technology. Many aspects of interviews can be automated, from scheduling to conducting the interview via video or chat, and scoring and recommending candidates. Interview automation can save time and money, reduce bias and enhance the candidate experience.
If you’re considering using interview automation, chances are it’s for one of these reasons:
It’s important to identify your reasons for using interview automation so that you can select the right technology, and set it up in a way that will help you achieve your goals. An interview automation process that prioritises saving money can look significantly different from one that prioritises candidate experience or bias reduction. (Although there are approaches that can achieve all of the above goals, easily.)
Take some time to rank the above reasons for using interview automation, and include any other reasons you have. Don’t be tempted to put increasing diversity in the too hard basket. As has been well-covered in other articles, having a diverse workforce isn’t just the right thing to do morally, it drives financial performance for your organisation.
A 2018 Boston Consulting study found that ‘companies with above-average total diversity, measured as the average of six dimensions of diversity (migration, industry, career path, gender, education, age), had both 19% points higher innovation revenues and 9% points higher EBIT margins, on average’.
New gadgets to automate interviews pop up seemingly every month. There are four main categories of interview automation technology. Interview automation has evolved from scheduling to asynchronous video and from Ai video assessment to Ai chat.
The first step towards interview automation was the automated interview scheduling technology. Most ATS (Applicant Tracking Systems) include interview scheduling features. Recruiters can shortlist candidates and send automated emails inviting candidates to schedule an interview online.
The next significant step in interview automation was the rise of asynchronous video. This allowed for candidates to respond to interview questions on video recording, which could be reviewed by a recruiter later. Whilst this helped resolve scheduling conflicts, the experience for candidates of speaking into a camera rather than to another person, and for recruiters making the time to review the videos isn’t ideal.
Ai video assessment technology not only records the asynchronous video of the candidate but uses Ai to score candidates. This saves recruiters a significant amount of time if they rely on the Ai scoring alone. Many recruiters still watch the videos, so this approach doesn’t always save more time than plain old asynchronous video. The candidate experience issues remain when using this approach—candidates are still alone talking into a camera, which doesn’t feel natural for most people. On top of the existing issues of asynchronous video, many video assessment tools aren’t clear about how candidates are assessed and don’t share that data with candidates. As a result, some candidates refuse to use Ai video assessments, and industry bodies are demanding more transparency.
There are also concerns around the accuracy of technology which uses facial cues to assess candidates. As technology journalist, Minda Zetlin opined in Inc.com in early 2018, “Someone whose relationship recently ended or who has, say, a toothache may display signs of unhappiness, anger, or confusion that have nothing to do with that person’s usual personality or fitness for a job. In a human job interview, you can choose to tell the interviewer if there’s something going on that may affect how you come across. You can say it in a video interview as well, but the AI might screen you out before a human ever gets to hear it.”
Ai chat technology differs from the other interview automation technologies in many ways. Firstly, Ai chat technology is employed at the applicant screening stage, rather than the traditional interview stage which happens after CVs have been screened. In fact, good Ai chat technology doesn’t even take in the data included on applicants’ CVs. Every applicant gets an interview and all of the information is provided by the applicant themselves in an Ai chat interview akin to having an SMS or messenger conversation. This gives applicants the opportunity to take as long as they need to answer each question, and removes many of the markers that create bias (name, gender, education etc) from the process.
Notably, most Ai chat technologies are clear about how the Ai works and conduct regular checks to understand how the technology is working to address bias. Some Ai chat technology, like PredictiveHire, automatically shares feedback with both successful and unsuccessful applicants, resulting in candidate satisfaction scores of 99% and over.
Generally speaking, as interview automation has evolved, the benefits have increased, for recruiters at least. Some candidates have been vocal about their dislike for asynchronous video and Ai being used without transparent explanation.
|Automation||Save time||Save money||Improve candidate selection||Improve candidate experience||Increase diversity|
|Ai video assessment||Yes||Yes||Maybe||No||Maybe|
Whilst most recruiters welcome the time savings that automating interviews delivers, some are concerned that more automation puts their own jobs at risk. There will always be a role for talented recruiters in finding and hiring exceptional candidates.
Kevin Wheeler’s prediction from back in 2017 stands. He said, “The usefulness of a recruiter will be in their ability to make a final decision or guide a hiring manager to make that decision. Influencing and listening skills as well as the ability to build relationships within the firm and outside it with potential hires will be more valuable than traditional recruiting skills.”
Once you’ve identified your goals and selected an automation technology approach it’s time to check that you’ve got the right measures in place to ensure you can capture baseline metrics and measure your success (or, room for improvement).
If your goal is to save time, you’ll be looking for an improvement on overall time to hire (which includes the time from sourcing to acceptance including offer and interview) and time to interview itself.
Make sure you know how much time hiring and interviewing is taking you before you implement a new system so that you can measure your success.
Saving money is a common reason for automation. Two measures which help understand financial savings are cost per hire and empty chair cost. Cost per hire is the sum of internal and external recruiting costs divided by the number of hires. The empty chair cost is the cost to the business of not having someone in a role. The cost of an empty chair in some roles is more obvious than in others. Missing a salesperson who would bring in $30,000 a month is a clear example. Conversely, in other roles, you may be able to leave a seat empty for a week and improve cash flow with only a small impact on productivity (although you wouldn’t want to leave it empty for too long and have it impact the workloads and engagement of other team members).
Of all the metrics, quality of hire is the trickiest to perfect. Some people measure the quality of hire by the number of hires who complete their probation, others by hiring manager feedback. Whatever measure you choose, ensure you’re consistent in the data you collect.
Two measures help us understand how we’re tracking on candidate experience: candidate satisfaction ratings and interview completion rate. Many technology platforms have candidate satisfaction features built-in where candidates answer either a net promoter score type question (how likely would you be to recommend this type of interview to a friend) or choose their sentiment about the interview from a list of emoji reactions.
Interview completion rate is a supplementary measure that helps you to understand how many people who start the interview complete it. If a lot of people start the interview but don’t complete it, the experience might not be as user friendly as it needs to be, or there may be a technical issue.
Some interview automation technologies can be extremely effective in reducing bias, which can increase workforce diversity. It’s important to understand the current state and any improvements to workforce diversity. The measures which interview automation technologies have direct impact on are diversity of candidate pool and diversity of recommended candidates. In addition to those measures, keep track of diversity of hires and diversity of retained employees so that you can identify whether recommended candidates are hired in line with recommendations, and whether hired candidates stay.
PHAI (PredictiveHire’s conversational chat Ai) can screen 100,000 applicants within hours and can scale to any number of applicants without causing any impact on candidate experience. If we conservatively assume that every recruiter spends seven hours each day screening, spends ten minutes reading each applicant’s CV, and phone screens 10% of applicants for 30 minutes each, it would take a team of five recruiters 476 working days to do the same amount of work. Phai interview automation is a huge 600 times faster. Those speed differentials compound when the numbers grow because humans can’t scale, but technology can. You can see what the scale of impact is when you look at the cost and time differential for 1,000 applicants and 100,000 applicants.
Interview automation can deliver some huge competitive advantages, saving you time and money whilst reducing bias in hiring. But it’s just as important to understand what interview automation can’t do.
Correcting bias in technology is far easier than correcting bias in humans. Unconscious bias training comes from a place of good intention, but it rarely moves the needle on workplace diversity. That’s why it’s important to use the measures of diversity above consistently and intervene if necessary.
Some interview automation technology does an exceptional job of reducing bias. But even with bias out of the equation, there are other aspects which impact workplace diversity. There are still real and measurable skills and competencies that a candidate needs to be successful. They might be related to communications, to technical ability, people might need a certain licence or to be able to speak a certain language. An applicant without the requisite attributes won’t be successful regardless of whether they’re from an overrepresented or underrepresented group—and they may have been disadvantaged in acquiring skills or competencies in life up until this moment. From that perspective, an unbiased interview can be determined as unfair.
In cases where candidates from underrepresented groups aren’t getting through the hiring process, some conversational Ai technology will allow you to see where those candidates are being held up, and you can make a decision as to whether those attributes are required of new hires, or if they can be acquired on the job. It’s this level of granularity that allows organisations to understand what’s blocking more diverse hiring and address it. Look for a technology platform that delivers multiple scores, and which can identify performance against those scores from particular cohorts.
Interview automation can help build a diverse workforce, but having a diverse workforce isn’t the same as having an inclusive workplace. In order to keep the employees you hire, you’ll need to ensure everyone feels like they belong.
As in a face to face interview, the best questions to ask reflect the key requirements of the role.
Here are our six tips for great automated interview questions.
One of the places interview automation really shines is in volume hiring. Two retail giants, Bunnings and Iceland have embraced Ai conversational chat interview automation, with fantastic results.
As Australia’s largest hardware retailer, Bunnings recruit for over 360 stores in Australia and New Zealand. Every year they receive application volumes in six figures for frontline roles. Most applicants are consumers—people who love the Bunnings brand. That makes the candidate experience critical.
Bunnings uses PredictiveHire Ai to automate interviews. Every candidate that applies for a Team Member Customer Service role in-store gets an interview.
The automated interview experience is really friendly. It’s a simple five-question conversational experience and candidates respond by texting on mobile in their own time. Every candidate, whether they’re successful or unsuccessful gets insights back, with a coaching tip.
The process is inclusive and gives everyone a chance at securing a role. The Ai doesn’t know any sensitive information like gender, age and race. It only analyses the text responses. Since go-live, the Ai has interviewed close on 6,500 candidates. Using PredictiveHire has reduced time invested in hiring by 90%.
Candidate sentiment is at an exceptional 99%. Here is one candidate’s experience which sums up very well the thousands of comments received: “I have never had an interview like this in my life and it was really good to be able to speak without fear of judgment and have the freedom to do so. The feedback is also great. This is a great way to interview people as it helps an individual to be themselves. The response back is written with a good sense of understanding and compassion. I don’t know if it is a human or a robot answering me, but if it is a robot then the technology is quite amazing.”
Iceland is a British food retailer with over 100 stores in the UK and global export business. In just four months, Iceland received over 500,000 applications, and like Bunnings, Iceland job applicants are often customers. They selected PredictiveHire Ai to automate interviews.
Iceland wanted to find a way that delivered a level of fairness and consistency around how applications were screened that enabled store managers to reduce that amount of time that they spent on recruitment. Candidate experience was also a priority.
Rather than sending standard ‘bulk’ responses to applications that say ‘if you haven’t heard anything from us in two weeks take it that you haven’t been successful’, Iceland wanted to share personalised feedback. They believe companies have a duty to help individuals to understand why they haven’t been successful and to help them to be successful in their next job application.
The automated feedback provided by PredictiveHire helps applicants understand where their strengths are and where their development needs are.
Here’s what one candidate says: “I enjoyed the interview. It makes me believe Iceland as a company are people carers and their staff are more than just employees.”
Yes! Sure, from time-to-time, the Ai used in interview automation and other Ai used in the hiring process gets poor media coverage. And we should certainly be vigilant to select technology that can be clearly explained and is reliable.
As interview automation technology, and Ai in HR technology generally, evolves, we need the space to experiment safely. William Tincup explores this in his interview with Editor in Chief of TechFunnel Danni White. “I’m starting to see some examples of some really cool things that people are trying, and some of it’s going to be a mass failure,” he explains.
“Amazon tried a different algorithm to get a better slate of candidates, more diverse—it backfired. The story pretty much from every outlet was that they screwed up. What we all missed is that they were trying. They were putting in the effort. If we say that is important. If we say diversity, inclusion, belonging, equity, equality is important, well then we have to actually do it. So, we missed on Amazon. We should have told the story from a different perspective: yeah they got that wrong but, just like a lot of experiments you start with a hypothesis. The answer is to make data-driven decisions.” William is right—Amazon was conducting an experiment, they never used that algorithm. It was one step towards finding an algorithm that will work.
Companies should use Ai to assess job candidates if it helps them achieve their goals and if that Ai is consistently checked for bias. For example, PredictiveHire doesn’t collect attributes which could attract bias, so that data is built using Namsor (www.namsor.com) in order to validate the effectiveness of the platform. Namsor takes names of applicants and derives gender and ethnicity. That data is used to understand diversity at each step of the interview process.
Equally, organisations should check the other metrics and measures of success regularly. They may need to adjust their approach: the questions they’re asking, the algorithms they’re using or the attributes they’re looking for, in candidates in order to achieve their goals.
Ai-based solutions are by far the most scalable solution and deliver far greater bias removal than other approaches.
Algorithms are more consistent, reliable and fixable than humans:
It is important that your interest in Ai bias is not limited to the algorithms. Ask also whether the method used to collect the data and how the human recruiters use the data. For example, a video interview can lead to bias whether it is being evaluated by humans or Ai. Lengthily psychometric testing on the other hand may give you unbiased results but at the expense of bad user experience and drop out.
The trick is to ensure humans use the data they’ve provided, as Tomas Chamorro-Premuzic and Reece Akhtar explain in Harvard Business Review: “Currently, many organizations that use digital interviews do not leverage these types of powerful AI analytics, as their recruiters are often unwilling to accept the algorithm’s recommendations and continue to rely on their own naïve judgement. Sadly, this ignorance is harming both the candidate and the organization. The HR departments that realize that science and data, and not intuition or instinct, should be the basis for decisions will attract and retain the best talent.”
So should companies use Ai to assess job candidates? The answer is yes. Of course, the Ai needs to be checked for bias, and we need to be transparent in how it works, but the benefits far outweigh the risks.
For you to achieve your goals using interview automation, it’s essential to set it up to get the most out of it.
5 steps to implementing interview automation
You can try out PredictiveHire’s FirstInterview right now – HERE! 😀
Artificial Intelligence (mostly Machine Learning) is being used more and more for high-impact decision making, so it is important to ensure these models are used in a fair manner. At PredictiveHire, we recognise the impact […]Read On
Competition for candidates today is fierce. COVID, border closures, BREXIT, the last two years have created a global candidate shortage that’s hitting large organisations hard. That’s why in today’s market, candidate experience is king. The […]Read On
This year, we have helped more than 60,000 people get a job. From retail assistants, graduates entering the workforce for the first time, sales professionals, team leaders and cabin crew, to name just a few. […]Read On
Get our insights newsletter to stay in the loop on how we are evolving PredictiveHire