1451. Mining Hidden Signals: Unstructured Data in SMME Performance Modeling
Invited abstract in session WA-23: Data Analytics for Business Resilience and Sustainability - Measuring SME Performance , stream OR for Societal Development.
Wednesday, 8:30-10:00Room: Esther Simpson 3.01
Authors (first author is the speaker)
| 1. | Helper Zhou
|
| School of Accounting, Economics and Finance, University of KwaZulu Natal | |
| 2. | Evelyn Zhou
|
| Financial Management, UNISA | |
| 3. | Gordon Dash
|
| Finance and Decision Sciences, University of Rhode Island | |
| 4. | Nina Kajiji
|
| Computer Science and Statistics, University of Rhode Island, and The NKD Group, Inc. | |
| 5. | Sharon Mandizha
|
| ENTREPRENEURIAL STUDIES AND MANAGEMENT, DURBAN UNIVERSITY OF TECHNOLOGY |
Abstract
The use of unstructured data in predicting Small, Medium, and Micro Enterprises (SMMEs) performance has been largely overlooked, with traditional models relying on structured financial and operational metrics. However, emerging studies—mainly in developed economies—suggest that qualitative factors, such as business descriptions and owner motivations, can offer critical insights into business success. This study bridges the gap by analyzing a dataset of 2,200 SMMEs across South Africa, leveraging sentiment analysis to assess how business narratives relate to financial performance. Findings reveal that owners who expressed highly positive sentiments about their businesses tended to be overly diversified and had moderate to low revenue. In contrast, those with more measured, neutral descriptions exhibited significantly higher revenue. This suggests that SMMEs may be influenced by optimism bias, where excessive positivity leads to overexpansion, diluted focus, and weaker financial outcomes. To mitigate this, policymakers should integrate sentiment analysis into SMME assessment frameworks and provide targeted training on strategic focus. Business support programs should incorporate structured self-reflection tools to help entrepreneurs assess their market positioning realistically, fostering more sustainable growth.
Keywords
- OR in Development
- Decision Analysis
- Analytics and Data Science
Status: accepted
Back to the list of papers