Home // DATA ANALYTICS 2016, The Fifth International Conference on Data Analytics // View article

Predicting Candidate Uptake For Online Vacancies

Authors:
Corné de Ruijt
Sandjai Bhulai
Han Rusman
Leon Willemsens

Keywords: recruitment analytics; HR analytics

Abstract:
The Internet has substantially changed how organizations market their vacancies and how job seekers look for a job. Although this has many benefits such as simplifying the communication, it can also cause problems. Some vacancies are obtaining more applications than can be handled by the recruitment department, while other vacancies may remain unfulfilled for a long time. Data analysis might reveal insights into what strategies are effective to solve these problems. To analyze these problems we therefore consider the predictability of the number of applications per vacancy per week, and to which extend this can be controlled using online marketing campaigns. After testing the predictive quality of several machine learning methods on a data set from a large Dutch organization we found that a Random Forest model gives the best predictions. Although these predictions provide insights into what recruiters and hiring managers can expect when publishing a vacancy, the error of these predictions can be quite large. Also, although the effect of online marketing campaigns on the number of applications is significant, predicting the effect from historic data causes problems due to collinearity and bias in the usage of these campaigns: a campaign is also a response to a small number of applicants who responded to the vacancy. Nevertheless, these predictions are insightful for both recruiters and hiring manager to manage their expectations when publishing a vacancy.

Pages: 63 to 68

Copyright: Copyright (c) IARIA, 2016

Publication date: October 9, 2016

Published in: conference

ISSN: 2308-4464

ISBN: 978-1-61208-510-4

Location: Venice, Italy

Dates: from October 9, 2016 to October 13, 2016