Differentiation in microenterprises

Abstract

Small unregistered firms contribute to a substantial proportion of global economic activity, particularly in developing regions. In explaining variation in productivity in these types of informal firms, research has focused primarily on the adoption of effective business practices and access to capital, with little focus on fundamental positioning. This article explores the nature of differentiation in microenterprises, introducing a text-based measure of differentiation using state-of-the-art sentence embeddings. Using a combined sample of nearly 10,000 microenterprises across eight developing countries, I examine whether (and which) microenterprises differentiate, whether differentiation is related to performance (and for whom), and whether any existing policy interventions affect differentiation. Managerial Summary: Small unregistered firms contribute to a substantial proportion of global economic activity, particularly in developing regions. In studying how to improve the productivity of these types of tiny firms, researchers and policy makers have primarily focused on business practices and access to capital, rather than the fundamental market position of the product or service offered. This article explores the extent to which microenterprises differentiate themselves from peer businesses, introducing a measure of differentiation based on computational text analysis.Using a combined sample of nearly 10,000 microenterprises across eight developing countries, I examine whether (and which) microenterprises differentiate,

Mobilizing Conceptual Spaces: How Word Embedding Models Can Inform Measurement and Theory Within Organization Science

“…Similarly, Guzman and Li (2023) use the doc2vec implementation of distance to measure the founding strategic differentiation of startups using the Crunchbase data. The authors find results similar to Carlson (2022), namely that new firms that differentiate experience increases in early-stage financing and equity outcomes.…”

“…As a result, scholars can train a conceptual space of technical innovation using the patent corpus, a conceptual space of public company communications using earnings call transcripts and press releases, a conceptual space of investment analysis using analyst reports, or a conceptual space of a specific firm's knowledge using internal communications such as slack and email messages. These conceptual spaces can be trained with minimal supervision and are therefore fast becoming valuable observatories for tracing the statics and dynamics of organizational life of central concern to organizational scientists (Hofstra et al 2020, Whalen et al 2020, Burtch et al 2021, Waller and Anderson 2021, Carlson 2022, Guzman and Li 2023, Lawson et al 2022, Lix et al 2022.…”

“…Others have used this measure to operationalize firm differentiation. In the context of microenterprises in developing countries, Carlson (2022) uses the BERT architecture (Devlin et al 2019, Reimers andGurevych 2020) to calculate the pairwise cosine distances for all microenterprises in their data set. With these distances, they are then able to estimate that across 10,000 microenterprises within eight developing countries, differentiation is associated with increased revenues and profits.…”

Renewing the resource‐based view: New contexts, new concepts, and new methods

“…A classic example is the use of various text analysis techniques, through which researchers convert text documents into quantitative measures. While early work in this area relied primarily on word counts based on pre‐defined dictionaries (Eggers & Kaplan, 2009; Uotila, Maula, Keil, & Zahra, 2009), scholars have increasingly used the power of AI algorithms to produce more sophisticated representations of unstructured text, such as the use of topic modeling (Choi, Menon, & Tabakovic, 2021; Kaplan & Vakili, 2015) and Word2Vec models (Carlson, 2022). These models have been applied to a variety of different sources of unstructured data, including letters to shareholders (Gamache, Neville, Bundy, & Short, 2020), earnings calls (Benton, Cobb, & Werner, 2022; Crilly, 2017; Guo, Sengul, & Yu, 2021), patent descriptions (Kaplan & Vakili, 2015; Kuhn, Younge, & Marco, 2020; Miric, Jia, & Huang, 2023), and online comments (Corritore, Goldberg, & Srivastava, 2020; Marchetti & Puranam, 2020), among others.…”

“…These techniques are valuable both because they enable researchers to study micro‐level concepts of attention and cognition in a granular way (Eggers & Kaplan, 2009; Guo et al, 2021) and because they allow for quantification and comparison of complex concepts such as strategy (Carlson, 2022), culture (Corritore et al, 2020; Srivastava et al, 2018), or communication style (Choudhury et al, 2019). Future work could continue to build on the increasing sophistication of these techniques, using them to represent and quantify other key strategy concepts; in particular, these methods might be used to measure intangible resources and capabilities that are hard to quantify using traditional methods.…”