Variables_of_Twitters_ brand_activity_influence_spreading_behavior

Published: 27 January 2017| Version 1 | DOI: 10.17632/2phnpkvz8z.1
Luis Matosas


Data The data shared shed light on the factors that mediate between brands and audiences in their relationships within a social media platform answering the two research questions presented on the article: RQ1: What are the constructs that represent audience spreading behavior of branded content within an online brand community? RQ2: Using brand controlled variables, what are the key indicators that predict the spreading behavior of branded content within an online brand community?


Steps to reproduce

Experimental Design, Materials and Methods A group of Twitter brands is selected from the population of companies in the Spanish food industry. The selection of food industry is based on the results of the Industrial Companies Survey carried out by the Spanish National Statistics Institute in 2015. The population is delimited to the code: CNAE 2009 - Manufacture of Food Products under the National Classification of Economic Activities taking firms with a turnover above 100 million Euros. The result is 147 companies, 50 of which show a brand presence on Twitter. After removing the inactive profiles in the last year, the final population is comprised of a total of 45 active brand accounts. In this final population of 45 Twitter accounts from the food industry, we gather all the content posted by our brands (Original Tweets, Genuine Replies and Retweets done) in a one year period from 4 February 2015 to 4t February 2016. The data collection for the research is performed using the Twitter API (Application Programing Interface) by means of its search and streaming functionalities. A total number of 41,392 pieces of brand-related content are retrieved. A two-step analytical approach is used. Firstly, exploratory factor analysis (EFA) is performed to cluster the dependent measures (expressed by Retweets and Favorites) generating the constructs for our structural model. Secondly, multiple regression analysis is conducted to identity and assesses the influence of the independent variables (those controlled by the brand) as predictors of brand audience response.


Marketing, Social Media, Consumer Behavior, Food Marketing, Social Media Marketing