How Algorithms Are Changing The Face of HR

Thanks to technology, the business world continues to move faster and faster. In this world of constant innovation, the most successful organizations are those that can keep up with or even get ahead of the technological trends. Just think of Google or Apple.

Of course, creating an innovative company is impossible without the right workforce. Hiring the right employees for the right roles at the right time is critical, and it requires good resource management to do it. That’s why many corporations are leaning more and more on HR data and algorithms to help with their decision making. In this article, I want to lift the lid on how some of those algorithms are changing the face of HR, and the kind of systems you can expect to see in your workplace within the next decade.

HR, Google, and People Operations

The rise of algorithm-driven HR can be traced back to Google. They were one of the first major companies to consider how to apply algorithms to human resource management. The company started by not calling it ‘HR’ in the first place. What most firms would call HR is called “people operations” at Google. It’s managed by the firm’s People Analytics Team and was spearheaded by Laszlo Bock in its early years.

Bock, the former senior vice president of People Operations at Google, wrote the best-selling Work Rules! in 2015, explaining the fundamentals of data-driven HR. That was before co-founding his current company, Humu in 2017. Humu creates and implements algorithms to provide insights into how to improve employee satisfaction, such as in the example below.

These algorithms and others in the field of HR gather data from work environments, surveys, CVs and any other data sources available to an organization. They then process and analyze that data to provide insights into things like boosting employee satisfaction, improving the hiring funnel, and more.

How Algorithms Are Changing HR

Driven by organizations like Google, algorithms are now being utilized across marketing and customer service in the form of conversation AI, and in many HR departments. You can compare it to how an organization uses Google Analytics to improve website performance, except that HR data focuses on employees and candidates.

One of the beauties of data-driven HR, or people analytics, is its flexibility. Algorithms can be adapted to measure specific elements which can help organizations address their most pressing HR concerns. These are the three areas of HR in which algorithms are having the biggest impact.

The Hiring Process

Any HR professional will tell you that their field isn’t only about hiring and firing. The fact remains, however, that overseeing hiring and onboarding is still key to HR. It’s also an area in which algorithms have a considerable influence.

The impact of hiring algorithms on HR has been widely tested and analyzed. A variety of different studies, including one by the National Bureau of Economic Research, have come to the conclusion that recruitment driven by data and algorithms leads to higher quality hires for companies.

The reasons for this, while diverse, boil down to a couple of major factors. One of the main benefits of data-driven decision making is the breadth of data sources available to algorithms. The second is the removal of human biases.

Algorithms can collect and process a huge volume of data. They can analyze information from CVs, publicly available information, and responses to assessments, which allows them to build a comprehensive picture of any and every candidate.

These data allow for the identification of qualities that make for a successful employee. Hiring teams can then look for those qualities within the skills and personality of candidates and choose people who are best suited for the job. Not only does this make for a better hire, but it also allows hiring teams to make decisions much faster than before.

The fairness of data-driven decisions also prevents human error from getting in the way of hiring the right candidates. Proponents of hiring algorithms argue that the algorithms can remove human biases—biases that color decision-making—from the equation. As explained by Nathan Kuncel, professor of psychology at the University of Minnesota:

‘We haven’t concluded that human judgments have no value. It’s just that these judgments come with a package that includes bias. People can get hung up on one piece of information and make too much of it.’

At best, biases can mean making the wrong hiring decision, and at worst, they can derail the hiring process completely. In theory, algorithms would remove that subjectivity.

With that said, the average person on the street would not like to have their CV screened by an algorithm. A study conducted by the Pew Research Center found that 76 percent of US workers would not want to apply for a job where their CV was screened by an algorithm, and most people think algorithms would do a worse job than a human.

There’s one final way in which algorithms have impacted hiring. That’s by analyzing and reassessing the hiring process itself. Part of Google’s early research into data-driven HR focused on the optimal length of the hiring process.

The results of the research led to Google’s so-called “rule of four” for interviews. They found that four interviews were optimal for hiring. Further assessment of candidates gave little additional value. Thus, Google recommended the shortening of what were often far longer hiring processes, a decision that saved both time and money.

Workforce Planning

Algorithms also shape HR when it comes to predictive modeling and analytics. Predictive modeling software uses algorithms to find patterns in large volumes of data, allowing people to more accurately predict future trends.

In the field of HR specifically, the data concerned would be information about an organization’s workforce. HR professionals and executives can use algorithms to identify factors that make for successful employees, that influence employee retention, and more.

Discovering what characteristics make successful employees can help your recruiting team find and attract better candidates for your organization. Meanwhile, identifying reasons for turnover and attrition is vital to workforce planning. Predictive modeling in this area can help firms answer the following questions:

• Who is at risk of leaving the company?
• What is it that persuades employees to go?
• What can we do to retain our best employees?

Answers to these questions help HR departments get ahead of any problems and take preemptive action to retain the staff they might otherwise lose. Such modeling can also help identify skills shortages and leadership needs, aiding firms in mitigating the risk of a resources gap.

In short, algorithms in this area allow for predictive workforce planning. Firms can identify potential issues or other patterns within their workforce, and they can do so before problems come to fruition. That allows them more time to find solutions.

Predictive modeling and analytics are particularly important to firms experiencing rapid growth. These are the businesses at most risk of workforce problems. When a company grows at pace, it’s more difficult to ensure that the workforce keeps up. Algorithms help these firms stay ahead of the curve.

Employee Satisfaction

Employee satisfaction and retention are key to business success. It’s how firms retain their edge over competitors. It’s also another of the major areas where algorithms are changing the face of HR.

As mentioned earlier, Laszlo Bock’s Humu is devoted to improving employee satisfaction. The way they do this combines data-driven HR and behavioral psychology. Data for the algorithms is drawn from a company’s work environment and internal surveys. They are then processed to identify critical behavioral changes that will have the biggest positive impact on the happiness of the workforce at large.

‘Nudges’ are then sent to leaders and managers within the firm by text or email. These ‘nudges’ suggest and recommend the prescribed changes. Humu has found that suggestions like these are more likely to be successful than direct orders.

Through machine learning, the process is then automatically tailored and improved. The algorithms monitor the outcomes of nudges, then tailor the timing and content of future messages based on those outcomes.

The best way to explain how Humu’s algorithms work in practice is with a case study. One such study was included in a New York Times article in late 2018, explaining how Sweetgreen, a salad chain, had used Humu’s algorithms.

The corporation’s main HR concern was employee retention. It costs much more money to recruit and train new hires than it does to retain current employees, not to mention that experienced workers are typically more productive and efficient.

In a company survey, 43 percent of workers admitted to having sometimes considered taking another job away from the company. Humu’s algorithms identified employee development as a chief concern. 81 percent agreed that Sweetgreen offered good development opportunities, and 88 percent expressed general happiness with the company. In other words, while most employees were generally happy with the company, there was still a significant number who were concerned about career development.

Store managers then received a ‘nudge’ suggesting they consider their team members’ career development options. Managers started to speak with individuals about the possibility of learning new skills and held meetings to discuss career pathways.

The Sweetgreen case study is a great example of how algorithms could apply in a practical sense. Data analytics can identify the cause of a problem (staff attrition in this case) and help you identify a solution. Then, you can start to implement that solution before your organization suffers negative consequences (like staff leaving).

Concerns

The path to innovation is seldom a smooth one. All paradigm-changing developments come with some complications, and that much is true of data-driven HR and algorithms as well.

Experts in the HR field have highlighted some moral and legal concerns, including unintended bias and questions of privacy. Laura Chapman, co-head of Freshfields’ people and reward practice in Asia, summarized these concerns well here.

Data-driven HR helps to remove human bias from recruitment, as mentioned earlier. However, poorly implemented algorithms could actually reinforce discrimination if they are designed to focus on the wrong character traits or qualities in applicants. In extreme circumstances, this can lead to certain groups being overlooked completely and thus an unfair hiring process.

The collection and use of employee data are what raises privacy concerns. Companies often rely on outside firms—like Humu—to collect and process that data. Because the security of personal information is very important to many people, some employees might have issues with sharing their data with such third-party companies.

Conclusion – The Future of HR

The concerns surrounding data-driven HR shouldn’t be minimized. However, that doesn’t mean algorithms in HR aren’t useful and valuable when used properly.

The success of Google, as one of the first to use data-driven HR, is compelling by itself. Add to that the studies which have provided evidence of its efficacy, and it seems data-driven HR is here to stay. By considering how HR algorithms (and other strategies involving algorithms) could fit within your organization, you can propel your organization into the future of HR and use the best tools and strategies to help your people at the same time.

About the Author

Phil Pearce is the founder of MeasureMinds Group, a Bristol-based digital analytics agency. He is an analytics expert, author, and web analyst. Over the past 15 years, he has been helping clients improve their analytics and search engine marketing through the introduction of new tools and disruptive techniques. He is renowned for his in-depth technical skills and the ability to solve business challenges through innovative technological solutions.