TY - JOUR
T1 - Antecedents and consequences of the key opinion leader status
T2 - an econometric and machine learning approach
AU - Ping, Yanni
AU - Hill, Chelsey
AU - Zhu, Yun
AU - Fresneda, Jorge
N1 - Publisher Copyright:
© 2022, The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature.
PY - 2023/9
Y1 - 2023/9
N2 - Key Opinion Leaders (KOLs) have an undeniable influence on businesses. Many online review communities, such as Yelp, give KOL users prominent status in their communities as cues of source trustworthiness. Using both econometric analysis and machine learning methods, we adopt an antecedents and consequences framework to investigate the drivers of KOL status and their economic impact on businesses. We find that a user’s social activity is more important in determining KOL status than the reviews themselves. On the consequences side, the paper shows that the first KOL review significantly boosts sales, regardless of the actual rating assigned by the KOL. After confirming this sales boost, we use random forest regression to predict sales using KOL review characteristics, including text. It is found that the number of KOL reviews as the most influential feature in predicting sales. This research contributes to the existing literature by adding a more granular, holistic investigation into KOLs in online consumer review communities.
AB - Key Opinion Leaders (KOLs) have an undeniable influence on businesses. Many online review communities, such as Yelp, give KOL users prominent status in their communities as cues of source trustworthiness. Using both econometric analysis and machine learning methods, we adopt an antecedents and consequences framework to investigate the drivers of KOL status and their economic impact on businesses. We find that a user’s social activity is more important in determining KOL status than the reviews themselves. On the consequences side, the paper shows that the first KOL review significantly boosts sales, regardless of the actual rating assigned by the KOL. After confirming this sales boost, we use random forest regression to predict sales using KOL review characteristics, including text. It is found that the number of KOL reviews as the most influential feature in predicting sales. This research contributes to the existing literature by adding a more granular, holistic investigation into KOLs in online consumer review communities.
KW - Difference-in-difference
KW - Key opinion leader
KW - Machine learning
KW - Online consumer reviews
KW - Reviewer certification
UR - http://www.scopus.com/inward/record.url?scp=85143586499&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85143586499&partnerID=8YFLogxK
U2 - 10.1007/s10660-022-09650-9
DO - 10.1007/s10660-022-09650-9
M3 - Article
AN - SCOPUS:85143586499
SN - 1389-5753
VL - 23
SP - 1459
EP - 1484
JO - Electronic Commerce Research
JF - Electronic Commerce Research
IS - 3
ER -