Handling volatility in application volume and internal mobility

In response to COVID-19, some employers have had to deal with rapidly expanded hiring demands requiring as much as 14 times their normal candidate volume. These spikes were most common in essential goods, retail, and logistics organizations. Others, such as retailers specializing in luxury or discretionary goods, childcare providers, and hospitality-sector organizations, implemented hiring freezes. In addition to pausing or substantially slowing down hiring, many of these organizations were required to furlough or lay off staff, including recruiters and other human capital management (HCM) professionals who assisted in creating or implementing the hiring strategy.

Very little empirical research exists to guide practitioners in how to optimally structure hiring practices during times of candidate scarcity as opposed to times of candidate abundance, whether those changes are short term (e.g., seasonal hiring fluctuations) or long term (e.g., economic recession). As such, talent acquisition and HCM professionals largely navigate such volatile hiring environments by relying on their own intuition and experience. Never has this been so apparent than during March 2020 when the unemployment rate changed from 4.4% to 14.7% (Morath, Reference Morath2020).

Many organizations are used to the cyclical nature of seasonal hiring. However, in our experience, seasonal hiring typically only increases volumes two to three times and, because the expansion is planned, can be staffed accordingly (e.g., hiring more recruiters or approving overtime). For some organizations, COVID-19 has resulted in both a substantially larger increase in the need for talent and one for which there was no plan. In one organization, operations panicked and proposed eliminating the use of a validated assessment process and instead hiring any applicant who met minimum qualifications. In another organization, leadership elected to keep their validated assessments in place but removed their interviews in an attempt to reduce time to fill.

In addition to the volatility in applicant volume prompted by the pandemic, many organizations have had to increase flexibility and mobility among existing talent. For instance, to avoid layoffs, some organizations have had to shuffle employees to other roles or reassign them to temporary roles. Some organizations are planning to extend these flexible work assignments beyond the pandemic. These changes in work structure present interesting questions for selection researchers and practitioners.

Accelerating digitization and automation of staffing processes

Some strategically oriented organizations did not need a pandemic to begin thinking about how new technologies can automate and streamline hiring. Such organizations were well situated to quickly increase their reliance on hiring technologies. As the economy transitions from a period of historically low unemployment to a forecasted labor market with high candidate-to-hire ratios, organizations must learn how to process a suddenly large number of candidates efficiently and fairly. Talent acquisition professionals and hiring managers are increasingly asking for technologies to facilitate, automate, and streamline as much of their hiring workflows as possible.

The organizations best equipped to leverage digital hiring technologies are already familiar with the methods and advantages of automating many of their processes. For example, standard screening questions meant to advance a candidate to next steps in the hiring process can be completed via text-based question and answer with a chat bot. Interview scheduling can be automated with technology that allows candidates to self-schedule (e.g., by syncing with interviewer work calendars). Artificial intelligence is now available to score open-ended responses (e.g., to interview questions) with as much accuracy as expert raters (Thompson & Mracek, Reference Thompson and Mracek2019). Continued research in validity, fairness, and candidate and employer perceptions are needed to help inform best practices in these exciting new applications of technology in virtual hiring.

Monitoring stability of test norms

Although many organizations stopped hiring overnight, others significantly expanded hiring as lockdowns were put in place across the world (Dickler, Reference Dickler2020). One finding is that previously stable norms often shifted significantly. This is noteworthy because (in)stability of norms can have implications for the utility and potential adverse impact of an assessment. For example, one assessment, normed on over 10,000 candidates, had a passing rate set at 70%. Shortly after COVID-19, candidates for this role significantly increased, and 45% of them were now failing the assessment. A cut rate of 45% is very aggressive, and the assessment was quickly renormed for the new population. Conversely, another organization found an increase in pass rates (from 75% to 82%) between February and April 2020. The opposite patterns observed across these two large employers presents interesting research questions regarding factors influencing changes in candidate quality before and during the pandemic.

Furthermore, if and when candidate pools return to those of prepandemic times, the test norms may shift back, and organizations may find they are screening out too few applicants. Organizations unable to constantly monitor and update norms or quickly adjust cut scores may expose themselves to increased legal risk without realizing it. For example, companies that stopped hiring during the pandemic may find when they resume, they have access to a very different, perhaps more skilled, pool of candidates who lost jobs to large-scale layoffs and furloughs. However, this scenario is speculative and needs to be researched and tested. Research seeking to identify antecedents of norm shifts due to changing candidate populations could help practitioners anticipate and adapt to such shifts. For example, researchers may be able to help inform employers’ understanding of the likely direction and magnitude of norm shifts based on changes in sourcing strategy, geography, and applicant pool populations.

Flexibility in interview type

One of the biggest changes in hiring has been a massive increase in digital interviewing. Figure 1 shows the minutes that our interviewing technology was used across all clients in April 2019. April 2020 showed a six-fold spike in minutes used relative to April 2019. Although digital interviewing is not a new technology, many firms were unprepared for social distancing guidelines that called for minimizing risk of COVID-19 exposure by avoiding unnecessary physical proximity of hiring managers and job candidates. Although in-person interviews may feel more normal to most interviewers, for the foreseeable future, they are likely to be less common. This raises a variety of potential questions—for example, Is it appropriate to allow for some candidates (e.g., those living locally) to be interviewed in person while others are interviewed virtually?” or “How do organizations best ensure candidates who are less technologically savvy or who do not have easy access to a smartphone or web camera are provided an equal opportunity?” Research suggests that interview modality can influence interview ratings and may also affect drop-out rates (e.g., Langer etal., Reference Langer, König and Krause2017; Van Iddekinge etal., Reference Van Iddekinge, Raymark, Roth and Payne2006). More research about mixed-modality hiring and digital interviewing will greatly help organizations as they determine how to implement these technologies.

Figure 1. Average Interview Technology Usage.

Note. Usage is scaled relative to average use from April 2019. For example, March 2020 showed approximately 3× the usage of April 2019.

Monitoring drivers of retention prediction

Although there is a large body of literature on generalizing the prediction of performance from one circumstance to another (e.g., Schmidt & Hunter, Reference Schmidt and Hunter1977), the prediction of retention is less stable across jobs, companies, and labor markets (Gibson etal., Reference Gibson, Koenig, Griffith and Hardy2019). In our experience, the prediction of turnover is inherently tied to the types of turnover most common for a specific role. Using a very basic framework of voluntary versus involuntary turnover, COVID-19 is likely to drive a substantial decrease in voluntary turnover for many roles as employees have fewer alternatives. But for some roles, we may see an increase in involuntary turnover, such as in companies that remove parts of their hiring process to hire faster, thereby hiring candidates not fit for the role.

Nonetheless, the answer is likely more complicated and nuanced than too many candidates per opening and hasty hiring decisions. As described by Rudolph etal. (Reference Rudolph, Allan, Clark, Hertel, Hirschi, Kunze, Shockley, Shoss, Sonnentag and Zacher2021), many frontline workers may be more at risk for increased job stressors, including larger workloads, more risks of exposure, and increased sense of job insecurity. The complicated and nuanced nature of likely drivers of turnover during and following the COVID-19 pandemic compared with times prior warrants additional study. Different types of turnover necessarily have different predictors, and if the nature of turnover substantially changes due to the pandemic, existing tenure predictors may need to be monitored and potentially adjusted if their efficacy is reduced. One example of a predictor that may be less effective in the future is biographical data (biodata) questions assessing traditional career stability. We have long known that biodata items of previous job tenure are related to turnover (e.g., Cascio, Reference Cascio1976), but many unemployed applicants may take the first job they are offered and continue to seek better employment. Research addressing how the COVID-19 outbreak and similar events contribute to the nature of turnover in organizations could help practitioners adjust their prediction models.

Predictors of success working from home

For the last few years, organizations have shown increased interest in whether predictors of success in a brick-and-mortar context are similar to those in a work-from-home context. Never have so many organizations been forced into such a prompt and widespread experiment of enabling a broad-scale remote workforce than right now. For instance, large technology employers such as Facebook and Twitter have made announcements of permanent moves to allow work-from-home arrangements (Conger, Reference Conger2020). Understandably, employers are now frequently asking whether what predicts success in an in-person environment also predicts success for those working remotely and whether assessments or their scoring should be adjusted.

For many employers, until recently, examining whether individual differences in high-volume roles were predictive of success in remote work settings was more challenging to evaluate. In traditional call center environments, for example, work-from-home opportunities may have been reserved for a small-scale pilot of high performers with a proven record of success in the role. Many call centers closely monitor their workers, tracking metrics and communications. Increasingly, remote monitoring and communication technologies enable comparable degrees of employee tracking and connectivity. Interesting research questions include developing a deeper understanding as to a theoretical basis for what differences, if any, we would expect to see in predictors of performance and retention for these different contexts.