TY - JOUR
T1 - Understanding value perceptions and propositions: A machine learning approach
AU - Kolomoyets, Yuliya
AU - Dickinger, Astrid
PY - 2023/1
Y1 - 2023/1
N2 - It is well established in marketing literature that aligning value creation with customers' aspirations promotes satisfaction, repurchasing, and competitiveness. This study employs structural topic modeling and sentiment analyses on online documents to provide an empirical account of value alignment from customers' and service providers' perspectives. This generates insights into i) the attributes valued by customers and service providers, respectively, ii) the valence of those attributes, iii) the sources of value formation, iv) the value alignment between customers and service providers, and v) the relative importance of value attributes for budget and upscale hotels. The results indicate that guests focus on interaction, cleanliness, and comfort, while service providers most frequently discuss service-related aspects; however, the first two attributes also affect hotel ratings. Furthermore, the sources of value differ in terms of valence. These insights show that structural topic modeling is a scalable approach to understanding value from both perspectives.
AB - It is well established in marketing literature that aligning value creation with customers' aspirations promotes satisfaction, repurchasing, and competitiveness. This study employs structural topic modeling and sentiment analyses on online documents to provide an empirical account of value alignment from customers' and service providers' perspectives. This generates insights into i) the attributes valued by customers and service providers, respectively, ii) the valence of those attributes, iii) the sources of value formation, iv) the value alignment between customers and service providers, and v) the relative importance of value attributes for budget and upscale hotels. The results indicate that guests focus on interaction, cleanliness, and comfort, while service providers most frequently discuss service-related aspects; however, the first two attributes also affect hotel ratings. Furthermore, the sources of value differ in terms of valence. These insights show that structural topic modeling is a scalable approach to understanding value from both perspectives.
KW - Structural topic modeling
KW - Value proposition
KW - Value-in-use
KW - machine learning
KW - Hotel attributes
U2 - 10.1016/j.jbusres.2022.113355
DO - 10.1016/j.jbusres.2022.113355
M3 - Article
SN - 0148-2963
VL - 154
JO - Journal of Business Research
JF - Journal of Business Research
IS - 113355
ER -