A Structural Topic Modeling Approach to Exploring E-Commerce and Online Startups during COVID-19

Abstract

Purpose: In this article, we investigate the emergence and success of start-ups in the e-commerce and online business sectors during the COVID-19 pandemic using structural topic modeling (STM). Composition/Logic: STM is a statistical technique that utilizes natural language processing (NLP) to identify patterns and trends in large text data. We analyzed business description text data from 7,933 start-ups founded between 2018 and 2021, sourced from the Crunchbase database. Findings: Our STM analysis identified four sectors—financial technology (fintech), educational technology (edtech), event planning, and social platforms—that we term the “COVID hot sectors,” which experienced significant growth during the pandemic. To further explore these findings, we conducted supplementary analyses using start-up funding data and the Google Community Mobility Report. Originality/Value: Our results suggest that the rise of the COVID hot sectors may have been influenced by social distancing and mobility restrictions during the pandemic and that start-ups in these sectors attracted increased attention from investors and stakeholders after the outbreak. This study has implications for understanding entrepreneurship during times of crisis and the business models of e-commerce and online start-ups.