By Mark Stone, Contributor
Recent findings about the benefits of a diverse and inclusive workforce confirm what we already know: Hiring a diverse workforce is not only the socially responsible thing to do; it is also smart business.
According to Deloitte, a diverse workforce is twice as likely to meet or exceed overall financial targets. Additional research from McKinsey & Co. finds that diverse companies are more innovative, profitable, and less susceptible to issues like groupthink. For example, companies in the top 25 percent for gender diversity in their executive suite are 15 percent more likely to experience above-average profitability than companies in the bottom 25 percent.
Yet despite mounting evidence of the benefits of diversity in the workplace, the reality is that ensuring a diverse and inclusive workplace is difficult, with issues to solve at every stage of the employee life cycle, from the resume screen to retention. One reason for this starts at recruitment as Dell’s Chief Diversity & Inclusion Officer Brian Reaves, explained – humans are often compelled to choose the applicants who resemble them.
“A desire to recognize and overcome bias is an important step forward,” he said. “But even with the best intentions, hiring decisions can still be influenced by personal factors we may not even be aware of.”
Fortunately, AI technology is gaining momentum in the recruitment field and is beginning to tackle some of the decisions that result in our unconscious bias. In fact, intelligent tools are popping up around the world to hire a more representative workforce.
Seattle-based Textio, an augmented platform that writes job listings, recently helped a large client increase the percentage of women recruits from 18 percent to 57 percent by using an intelligent text editor to analyze and suggest modifications to the job listing’s word choice to avoid gendered phrasing. Mya Systems, a San Francisco-based AI company, developed an intelligent chatbot to “interview” and evaluate job candidates; programmed to ask objective, performance-based questions, the idea is that Mya will help avoid the unconscious bias that could arise with a recruiter.
And emerging as a leader in AI recruitment is Ideal, an AI software company based in Toronto, that was founded in 2014 to combat the compulsion to hire those too much like ourselves.
“A desire to recognize and overcome bias is an important step forward. But even with the best intentions, hiring decisions can still be influenced by personal factors we may not even be aware of.”
— Brian Reaves, Chief Diversity & Inclusion Officer, Dell
Sometime around 2012, Ideal co-founder Shaun Ricci realized that the recruiting process he had in place needed improvement. While his former company was able to hire around two salespeople per week, they consistently faced a high turnover rate.
“Looking back, we had no scientific way of recruiting,” he said. “We’d look at a resume, and based on a gut feeling, we’d decide if they’d make a good candidate. We made all the classic mistakes.” As a result, Ricci and co-founder Somen Mondal set out to create an AI system that would behave like a virtual assistant and apply it to specific recruiting tasks.
Today, Ideal’s software connects to a customer’s application tracking system to screen prior candidates as well as to external databases (such as CareerBuilder) to automatically source candidates for an open position. It then crunches both historical and current data to paint a picture of the qualities a strong candidate should possess, producing a “report card” with information such as job fit, skills match, resume quality, and a percentile score in comparison to other candidates.
Ricci explained that the software also looks at external inputs such as employer assessments and whether the candidate was a top performer in previous roles.
“We feed all that data into our machine learning data model for what a good candidate looks like for a specific role at a specific company,” Ricci said. “The next time a new candidate applies, we compare it to the model, we screen the resume, and if there’s any data that’s not on the resume, we can connect with them via chatbot to ask additional questions.”
Ideal’s goal is to free up recruitment teams so they are able to do the important tasks, such as building relationships with candidates. “There’s no reason a person should spend eight hours reading resumes,” Ricci said. “We have self-driving cars—we should be able to tackle resume reading.”
Time savings notwithstanding, promoting diversity is an important element of Ideal’s AI software. Ricci said that one key variable in mitigating unconscious bias is the ability for the algorithm to exclude the candidate’s name during the interview process. “There’s a lot of research on this; an English sounding name on a resume versus an ethnic sounding name on the exact same resume has been shown to have a 40 percent higher likelihood for being selected for an interview,” he said. Scrubbing names—the default setting of the AI algorithm—is one measure Ideal takes to help companies avoid unconsciously favoring any one demographic.
Another measure Ideal relies on to promote diversity is employing an industrial organizational psychologist whose role is to study psychology in the workplace. With the psychologist on staff, Ideal aims to eliminate mimicking human biases in the human-developed AI system. Ricci explained that the psychologist does extensive research on how bias creeps into the hiring process and “teaches” the algorithm to how to avoid this bias.
The Employee Experience Frontier
While today’s AI solutions can help propel the right candidate to the interview stage, biases can creep back into the picture at any stage of the employee experience.
As artificial intelligence moves forward, experts predict different types of technologies will tackle biases at all stages of recruitment. “Emerging technologies like augmented or virtual reality can help prevent bias in interviews,” Reaves said. “Through the technology filter, visual and auditory clues about the candidate can be obscured, enabling companies to make hiring decisions solely based upon what the candidate is saying, not how they look or sound.”
“Through the technology filter, visual and auditory clues about the candidate can be obscured, enabling companies to make hiring decisions solely based upon what the candidate is saying, not how they look or sound.”
But even if a company is successful in hiring a more diverse workforce, can they ensure diversity and inclusion thrive as employees continue their career? This next step may be the next big thing to watch for in AI technology.
“While recruitment is a current focus to alleviate bias, as companies continue to implement technologies, they can begin to track a candidate’s entire journey through the company,” Reaves said. “This will provide insight into all stages of the career life cycle, and identify additional areas where bias may be impacting a person’s day-to-day responsibilities and potential for advancement.”