High Volume Recruitment: Success through Analytics

Many organizations recruit a large number of people in a short amount of time, a practice known as high volume recruitment. Organizations hire in volume for various reasons. For example, they are growing exponentially, they are opening new locations or offices, or to create economy of scale and synchronize with the school calendar, they have regular recruiting seasons during which they hire in volume. In a recent paper, Professors Lilun Du and Qing Li of HKUST show how this often-challenging form of recruitment can best be managed. “To ensure success,” they say, “three elements are critical: a simple and consistent process, data, and analytical tools.” Addressing all three elements makes this “the only comprehensive study that shows the what and the how” of high-volume recruitment.

Mass recruitment is difficult for several reasons. “Because of high volume,” the researchers explain, “the amount of resources allocated to learning about each candidate is small.” Therefore, organizations are often forced to choose applicants based on insufficient information. Compounding this problem, add the researchers, the “time constraints and large number of people involved mean that each recruitment decision must be made very quickly.”

To facilitate the decision-making process, the authors tell us, “many recruiters use Internet-based systems for accepting applications, evaluating candidates, and sending out offers. Data are readily available, relatively clean …” This makes high-volume recruitment a strong candidate for the use of data analytics, which can “streamline the recruitment process, save time, and allocate limited resources to where they are most needed.”

Seeking to “help recruiters make more informed decisions with respect to who should be short-listed and accepted,” Professors Du and Li proposed a novel data-driven approach to high-volume recruitment. In their model, the recruitment process consists of two stages: screening and interview. All candidates are evaluated in the screening stage, but only those who with sufficiently high screening scores are short-listed for an interview. After the interview stage, offers are made based on the screening and interview scores. The model provides tools to determine the number of candidates that should be short-listed and how the interviewed candidates should be ranked based on the two scores, in order to ensure the probability of hiring the wrong candidate or rejecting the right candidate at an acceptable level,

To test their model, the authors examined real-life data from admissions to a postgraduate business studies program across four academic cohorts. One interesting finding from that case study is that the interview score has more predictive power than the screening score.

The model developed by the authors can help program administrators in universities, human resources professionals in businesses, and recruiting companies to leverage data analytics for high-volume recruitment. The approach “is applicable to any situation in which an organization recruits people in batches and the recruitment process involves multiple rounds,” say the researchers. They hope that the study “will stimulate more interest in the use of analytics in recruitment in general.”


Assistant Professor
Information Systems, Business Statistics & Operations Management


Academic Director, MSc in Global Operations, Professor
Information Systems, Business Statistics & Operations Management