Haig R. Nalbantian, the dean of Workforce Science, discusses about the role of Technology in workforce strategy and management. Including the implications on recruiting and social networks.
Prometheus: What is the outlook for Technology’s impact on the workforce?
Haig: I must say, when all the chatter about the future of work started almost a decade ago, I was quite skeptical of the dire predictions coming from academic and management consulting circles concerning the prospect of computerization and AI displacing large percentages of then current jobs. At the time, I just wasn’t seeing such dramatic change among our clients. In fact, looking over the last decade, the actual trajectory of job displacement has been significantly slower than anticipated. That said, there is no question our clients are thinking very seriously about the impact of technological change on their workforces. This is not an abstraction for them, this is not a futurist exercise. Indeed, for many organizations, the future of work is already here and Covid-19 is only accelerating its arrival.
Companies are very eager to understand how new technologies infiltrating the workplace, including digitization, AI and automation, affect the talent and organizational requirements of their business. They want to determine if and how they can adapt their current workforces to the new requirements or whether they will need to bring in new and different kinds of talent than they have traditionally employed. And they know they need to assess to what extent traditional employment — even “the job” — will remain the dominant workforce model or whether a more task-driven approach to work will require greater reliance on market-based, labor transactions as opposed to regular, full-time employment.
These are big questions to address. It’s clear there is no escaping them. So, anyone focused on workforce strategy, management and planning is obliged to get his or her hands around the impact of technology. In my own work with clients, anticipating the talent implications of new technologies has become a central consideration. And it’s not just a consideration for workforce planning; it has ramifications for strategies related to diversity, equity and inclusions as well, offering new opportunities for organizations to further diversify their workforces.
One technology-related area in which I do have significant experience concerns the role of information in driving better workforce decisions. For many years I and my colleagues at Mercer have used workforce and business performance data managed by IT to help clients inform their decisions about workforce management, through the application of advanced analytics, such as internal labor market (ILM) modeling®. Using this approach, they have learned a lot about the workforce characteristics and management practices that drive value from their workforce.
They also learn some very specific things such as what drives employee turnover or what are the core factors predicting career success. These predictive models have many practical applications. For instance, turnover and success profiles can be used to help on the recruiting front. They can be the basis of predictive hiring methods that help improve the quality of hires. Also, these modeling results can be used to give potential employees the opportunity to view what a career looks like in the organization and see how and where they would fit. This can be quite exciting for job candidates.
Prometheus: What about the implications for recruiting — or ‘predictive hiring?’
Haig: Some major employers like IBM, Citigroup, JetBlue, to name a few, deploy technology-enabled tools to support talent acquisition. These tools fall under the banner of “predictive hiring.” Many of these predictive hiring tools are based on what we call “matching” models that analytically align the experiential and behavioral characteristics as well as preferences of job prospects with those required to be successful on the job and/or with the employer. Some involve real-time screening of attributes that may not show up on a resume, for example, gamification in which the employer gains visibility about the way the job prospect approaches a problem, reacts to information or unexpected events and/or eventually solves the problem or completes the task.
This can be quite powerful. First, to the extent the matching mechanisms and/or predictive models are reliable, they help close the information gaps that exist between the buyers and sellers in labor markets, ultimately leading to better matches. In a word, they help illuminate behavioral traits that are otherwise unobservable pre-hire and that might not even be exposed through formal testing. Second, predictive hiring models can expand opportunity for employer and prospective employees alike by bringing in candidates who might otherwise have been ignored because they didn’t precisely match the job specifications. This enables the employer to look more broadly at the capabilities and attributes of people they consider and maybe trade off pedigree for certain observed behavioral proclivities.
But there are serious challenges to predictive hiring as well. On the downside, two things really jump out. Number one, the methods themselves may select out people in certain income or demographic groups who may not have the same experience with technology to be able to use it as well as others. For example, older workers are likely not as well versed in gamification. They didn’t grow up on the computer playing video games. Likewise, candidates from low-income communities with more limited access to computers and training, may be at the same disadvantage. There is a very real risk of introducing a new form of systemic bias into the recruitment and selection process that can easily rival subjective assessments.
The other thing I’ll say, just based on my experience with these tools, is that they are best suited to fill jobs with easily measured individual performance outcomes. So, if you were just looking at the proficiency of, say, a product sales professional, you might make the case — I need a set of clones who match the success profile. But in many environments, performance is driven by complementarities among team members. Performance reflects group capability and dynamics, not the sum of the individual parts. Having people who complement each other’s skills and experience is more conducive to performance than having a team of people starkly similar in background and skills. Building a predictive hiring mechanism based on group requirements is more complex than matching to an individual job or role.
This is personal to me. I can tell you in our own work, trying to identify complementarities, matching clusters of skills, and behavioral competencies that drive better team outcomes, instead of just identifying individual characteristics or behaviors associated with better performance in a particular role, is a challenge, but one well worth the effort. We would not be successful, in my view, as a team of substitutable clones, all with the same kind of analytical skills. We are effective precisely because we are a team of highly diverse professionals with complementary backgrounds, aptitudes, skills and capabilities. I don’t think a predictive hiring algorithm would serve us well.
I fear excessive reliance on these technology-driven recruiting methods may undermine the performance of teams. And since ultimately the performance that is most important to our clients is team or organizational performance, this is no small consideration.
Prometheus: How has exploring social networks improved diversity and enabled better team performance?
Haig: Our team does a lot of work applying advanced analytics to inform Diversity, Equity and Inclusion (DEI) strategies of client firms. We find in that work that relationships at work — e.g., whom you report to, the characteristics of your team, the customers or clients with whom you work, etc. — matter a lot in predicting what will happen to you from a career and performance perspective. One way to better understand these relationships is the application of Organizational Network Analysis (ONA).
This methodology helps explore employees’ social networks at work, the set of people with whom an individual interacts routinely in the course of his or her work. For example, what position do you occupy within your network of interactions? Are you in a central position or do you sit on the fringe? How close are you to those who are in a central position? To what extent are you a bridge to other social networks at work? How much work-related information passes through you? How frequently are people coming to you with questions and seeking help? Using ONA, organizations can gain a perspective on the nature of employee networks within, shedding light on many different issues of policy relevance.
For example, ONA can help uncover differences between formal and informal hierarchical structures, identifying actual leaders at all levels, and determining if and how demographics influence the position of employees within social networks. Regarding the latter, we have examples where we found that minorities are less likely to have high “centrality” within their networks. They are more likely to be on the fringe of their networks which means that compared to their white counterparts they’re neither getting nor conveying substantial business-relevant information on a regular basis, This may help explain why they’re less likely to get into the “right” roles to begin with, and why when they’re in the right roles they may not benefit as much from it because they’re still marginalized from a group dynamic point of view.
So much insight can emerge from objective assessments of employee networks. When organizations uncover demographic disparities in network patterns, there’s high motivation to do something about it because it’s so clear how they can cascade through the system to impede the performance and career progress of those affected. Addressing harmful imbalances in networks requires the active engagement of employees along with leadership. It’s not something that can be accomplished unilaterally through policy changes.
Yes, an employee can undertake efforts to expand his or her networks and become more central to the network(s) of which they are part. But reciprocal efforts by others in the network are required. This process takes on the characteristics of what game theorists call a “cooperative game.” In my experience, engineering a transition from a “non-cooperative game” context to a cooperative one is notoriously hard to accomplish and then sustain in organizations. But it can be very productive if achieved. Significant culture change is often required to make that happen. Fortunately, many organizations are starting to realize this, so I am optimistic that significant progress will be made.
Haig R. Nalbantian, Senior Partner and co-founder/co-leader of Mercer Workforce Sciences Institute, bridges the world of data analytics and creative forecasting as companies plan for the future. Haig has been instrumental in developing a “new science” of human capital management — “Workforce Sciences.” He and the Mercer team analyze workforce and business performance data to measure and identify where businesses derive value from their human capital and determine which management practices effectively increase this value.
