Detecting customer queue "at-risk" behaviors based on ethograms to minimize overall service dissatisfaction
Chapitre de livre
ICSOC : International Conference on Service-Oriented Computing Service-Oriented Computing – ICSOC 2017 Workshops : ASOCA, ISyCC, WESOACS, and Satellite Events, Málaga, Spain, November 13–16, 2017, Revised Selected Papers
(pp. 18-29). 2018,
Cham : Springer
Lien vers la publication
Every service encounter corresponds to a “queue network” in which a system of waiting lines is connected to servers. We posit that each production service type (e.g., restaurant, airport) requires an adapted queue design in order to maximize attributes salient to customers (i.e., their primary elements of per-ceived value) in today’s globalized service environment. While the queues have been studied from many angles, a scientific contribution based on a human ethol-ogy approach proposing the early identification of “at-risk” behaviors to regulate queue dynamics seems to be novel. To remedy this shortcoming, the large-scale food distribution sector has been chosen for the application of a naturalistic ob-servation approach to describe in detail the behavior of customer queues. Sixteen immersion episodes were conducted in the months between May and June 2016. Using RQDA, we analyzed the immersion transcripts and identified typical cus-tomer queue behavioral patterns. Then, we developed an ethogram containing what we considered to be “at-risk” queue behaviors. This ethogram can ulti-mately be used as an anticipatory indicator in the context of feedforward man-agement controls. Feedforward control, as opposed to classical feedback con-trols, is based on the early detection of risks and the implementation of mitigation before damage occurs. While this approach requires human attention and exper-tise (which can be quite expensive and labor-intensive), there is also potential for human ethology to assist managers with supportive or complementary automa-tion. Indeed, the factual description of behaviors contained in our ethogram can easily be coded with modern technology like facial expression and body recog-nition technologies.