ttllogo.png

People, Person, Computer, Electronics, LCD Screen, Laptop, Pc

TTL LookBack V3: Can Big Data and Machines Predict Successful Hires Better than You?

Welcome to Talent Tech Labs Lookback. We spend a lot of time exploring the Talent Acquisition Ecosystem and Marketplace and while every TTL Trends Report focuses on a different theme, we occasionally like to look back at some of our greatest hits.

Predictive Analytics for Hiring: Can Big Data and Machines Predict Successful Hires Better than You?

Rapidly evolving digital information technology has opened the door for game-changing HR innovation, providing new and practical solutions for HR success in key areas that have always been especially challenging—recruiting and hiring.

This innovation is vital: Organizations struggle to fill jobs on a global basis despite stubbornly high unemployment. More than 201 million people worldwide are unemployed today, according to the International Labour Organization, and that’s not counting those who have dropped out of the labor force. There’s a mismatch between the roles that organizations need to fill and the talent that is available in the global pool. While an aging workforce and a skills gap contribute to this mismatch, could outdated hiring practices also share the blame?

In our dynamically networked world of social media, cloud and crowdsourcing, plus big data, there is exponential growth in the volume and variety of data that companies can rely on to analyze workforce needs and recruit employees. HR leaders must embrace new testing and data strategies as they fast forward through the first quarter of our wired century to fresh horizons of talent assessment, acquisition, development and mobility.


Predictive Analytics for Hiring: Take a look back at what @Mercer predicted in 2015 in our Trends Report V3!
Click To Tweet


Predictive hiring: how it works

Recruiters and hiring managers have always worked to identify and hire the candidates they predict will be successful at their companies. Some trust their instincts while others use screening and assessment tools. What’s different now is that big data and machine-learning algorithms can provide quantitative information to help make that prediction.

Instead of deciding what skills and competencies are needed for a role, big data can zero in on the traits and attributes of high-performing employees. With the data telling the story, it’s easy to define the profile of top performers. With that profile data, machine-learning algorithms can screen through hundreds, thousands, even millions of candidates to find and flag for further evaluation those that match the profile. Even better, the machines get smarter and smarter as more data runs through them.

Man vs. machine

“Humans are very good at specifying what’s needed for a position and eliciting information from candidates—but they’re very bad at weighing the results,” according to an article from Harvard Business Review (HBR). HBR’s analysis of 17 studies showed that “a simple equation outperforms human decisions by at least 25%.”

There are many reasons why algorithms are able to beat human instinct in hiring—for example, according to HBR, people are easily distracted by irrelevant information. Hiring managers inevitably have conscious and unconscious biases; even with the best of intentions, they can’t help but select people who think and act like they do.

Machines win, but algorithm aversion takes over

If big data and machine algorithms are proven to be better at hiring than humans are, why aren’t these big data/predictive hiring approaches being adopted faster?

In a 2014 study published in the Journal of Experimental Psychology, research showed that evidence-based algorithms more accurately predict the future than do humans. Yet when forecasters are deciding whether to use a human forecaster or an algorithm, they often choose the human forecaster.

This phenomenon, called algorithm aversion, shows that people will often be averse to algorithmic forecasters after seeing them perform, even when they see them outperform a human forecaster. In short, we are quick to forgive ourselves when our instincts fail us—and, for example, we hire the wrong person—but we still don’t trust algorithms to make the predictions for us.

Moving to a pro-candidate solution

About 60 to 70% of companies use selection assessments to predict successful hires. These assessments, which may examine technical or non-technical knowledge, skills and attributes, prove helpful to hiring managers but are often candidate-unfriendly. As a result, many potentially qualified candidates drop out of the recruitment process. Millennials in particular express frustration with or are unwilling to complete traditional hiring assessments.

However, fun and engaging solutions are being designed to use big data and machine-learning algorithms to assess the behavioral skills of entry- through mid-level managers. Innovative approaches such as “serious gaming” and “gamification” assess candidates via mobile and other platforms. This works well for a new generation of talent that thrives on real-time competition and immediate feedback.

Serious business for HR, serious fun for candidates

With gamification, candidates compete with peers, and hiring managers gain valuable insight and information revealing candidates who may have the right job skills but could have been overlooked using traditional credentialing or sourcing.

Candidates can be immersed in solving the sort of problems they will face as an employee or as a manager using on-the-job simulations that combine narration, video, photos and animation. Simulations showcase a candidate’s problem-solving abilities, not by asking them what problems they’ve solved in the past, rather by engaging them in actual problem solving. In another scenario, candidates can play neuroscience games that are grounded in decades of academic research. This approach not only helps candidates determine which jobs and careers they may be best suited for but also allows companies to see if candidates match the cognitive and emotional trait profile of their most successful employees. These new game-like approaches test candidates’ skills and knowledge, managerial potential and job preferences. At the same time, it can be a fun and informative process for candidates and give them greater visibility into their potential employers, well beyond conventional job board and resume-dumping sites.

Visualize a sustainable network of worldwide talent The ability to connect to and prequalify the multitude of candidates around the globe is now here via social networks, big data and machine algorithms. Companies that embrace these new talent acquisition innovations will have an early adopter advantage. It will take vision and partnership to manage these innovative new frontiers, but the payoff—a sustainable network of worldwide talent—will be worth the effort.

About the Author: Barbara Marder is a member of the Talent Global Executive Leadership team, and is the Global Innovation Leader at Mercer. In this new role, Barb oversees an innovation team and creates new offerings for Mercer’s talent business. She is leading the innovation team from idea generation to testing key assumptions, to commercialization on the new offering.

The post TTL LookBack V3: Can Big Data and Machines Predict Successful Hires Better than You? appeared first on Talent Tech Labs.

Read more >

Published on Jan 29, 2018

ttllogowhite.png