Journal of
Marketing Development and Competitiveness






Scholar Gateway


Abstracts prior to volume 5(1) have been archived!

Issue 5(1), October 2010 -- Paper Abstracts
Girard  (p. 9-22)
Cooper (p. 23-32)
Kunz-Osborne (p. 33-41)
Coulmas-Law (p.42-46)
Stasio (p. 47-56)
Albert-Valette-Florence (p.57-63)
Zhang-Rauch (p. 64-70)
Alam-Yasin (p. 71-78)
Mattare-Monahan-Shah (p. 79-94)
Nonis-Hudson-Hunt (p. 95-106) 



JOURNAL OF APPLIED BUSINESS AND ECONOMICS


Predicting Amazon’s Choice of HQ2 from Social Media: Evidence From the Tweets of Informed Sources


Author(s): Kissan Joseph, Abir Mandal, Sumanta Singha

Citation: Kissan Joseph, Abir Mandal, Sumanta Singha, (2020) "Predicting Amazon’s Choice of HQ2 from Social Media: Evidence From the Tweets of Informed Sources," Journal of Applied Business and Economics, Vol. 22, Iss.10,  pp. 214-231

Article Type: Research paper

Publisher: North American Business Press

​Abstract:

Social media chatter, and in particular, Twitter, is increasingly gaining popularity to generate forecasts in a wide variety of domains. We build on this body of work and set out to predict Amazon’s HQ2 choice by analyzing the tweets of officials at the 20 finalist cities. Consistent with the affect infusion model (AIM) from the psychology literature, we conceptualize that the positive affect generated in successful ongoing negotiations will lead to a congruent positive spill over even in unrelated tweets. Analyzing tweet series that include a corpus of 50,238 tweets and incorporating dynamic time warping measures, our forecasting method correctly predicts Northern Virginia, favors it over two proximal cities, Washington D.C. and Baltimore, and ranks New York City 11th out of 20 cities. These forecasts match those of the betting markets. Our research thus offers an alternate and novel approach to extracting the signal from the noise in social media.