Defining Brand Perception: Social Media Data Analysis of Food Delivery Services

Purpose – This paper aims to define problematic areas of the food delivery industry from a customer perspective by analyzing Twitter data and to provide up-to date recommendations based on the framework of the AIDA model. Design/methodology/approach – Five food delivery services company, i.e. UberEats, Just Eat UK, Doordash, Pizza Hut, Dominos, were considered in this study. Twitter was the data collection source. Data was scraped using python's Twint package for Q4 2019 and Q4 2020 to investigate whether the pandemic affected the food tech industry or not. Research design contains sentiment analysis on sentence and a word level for each brand, emotional valence for the whole food delivery industry, word frequency analysis through pairwise correlation and topic modeling as a tool for forming brand related recommendations. Findings — User generated content in Twitter on the selected delivery service companies has both food tech industry-specific similarities and differences. The study contains recommendations for each brand with the information what specific feature of the product or service companies’ management should improve. Originality/value – The relevance of the research topic proved by the fact that the previously proposed list of food delivery companies was not considered together. Also, selected companies were not studied in the context of world changes caused by the pandemic. Scope of analysis is broadened by a combination of such methodological approaches as topic modeling based on sentiment analysis. Keywords: Food delivery service, FoodTech, Sentiment analysis, Topic modeling, Text Analysis Paper Category: Master thesis