Using Social Media and Personality Predictions to Anticipate Startup Success
(2023) In Master's Theses in Mathematical Sciences FMSM01 20232Mathematical Statistics
- Abstract
- This thesis explores the potential of integrating predicted founder personalities, based on the Big 5 Personality Framework, into Machine Learning (ML) models to enhance the accuracy of early-stage startup success predictions. Leveraging Natural Language Processing (NLP) techniques, we extracted personality insights from founders' tweets, focusing on US startups funded between 2013 and 2015. Our research utilized a range of models, including XGBoost, Random Forest, and Feed-forward Neural Network for personality predictions, and Logistic Regression, XGBoost, and Random Forest for startup success forecasts. Results indicated that most personality-predicting models outperformed the Naive baseline. In success predictions, XGBoost emerged as... (More)
- This thesis explores the potential of integrating predicted founder personalities, based on the Big 5 Personality Framework, into Machine Learning (ML) models to enhance the accuracy of early-stage startup success predictions. Leveraging Natural Language Processing (NLP) techniques, we extracted personality insights from founders' tweets, focusing on US startups funded between 2013 and 2015. Our research utilized a range of models, including XGBoost, Random Forest, and Feed-forward Neural Network for personality predictions, and Logistic Regression, XGBoost, and Random Forest for startup success forecasts. Results indicated that most personality-predicting models outperformed the Naive baseline. In success predictions, XGBoost emerged as the top performer, showcasing the highest scores in Macro F1 and AUC for both Series B and Series C funding rounds. While the trait of Neuroticism was highlighted as significant for Series B predictions across models, Series C predictions emphasized the importance of Openness and Agreeableness. Our findings underline the value of integrating predicted personality traits into ML models for startup success forecasts. However, as with all research, our work had inherent limitations and suggested areas for further exploration and improvement. (Less)
- Popular Abstract
- Decoding Startup Success: How Founder Personalities Shape the Future
Predicting the success of startups is notoriously challenging, partly due to the significant information asymmetry in the entrepreneurial world. My research delves into this intricate landscape, investigating a compelling question: can the personalities of startup founders, as seen through their social media presence, offer insights into the future of their ventures?
At the heart of this study is the 'Big 5' personality framework – openness, conscientiousness, extroversion, agreeableness, and neuroticism. These traits, fundamental to our behavior and decision-making, could potentially influence a startup's trajectory. I embarked on an analytical journey to uncover... (More) - Decoding Startup Success: How Founder Personalities Shape the Future
Predicting the success of startups is notoriously challenging, partly due to the significant information asymmetry in the entrepreneurial world. My research delves into this intricate landscape, investigating a compelling question: can the personalities of startup founders, as seen through their social media presence, offer insights into the future of their ventures?
At the heart of this study is the 'Big 5' personality framework – openness, conscientiousness, extroversion, agreeableness, and neuroticism. These traits, fundamental to our behavior and decision-making, could potentially influence a startup's trajectory. I embarked on an analytical journey to uncover these links, turning to the digital narratives of U.S. startup founders on Twitter between 2013 and 2015.
Using machine learning methods, a branch of artificial intelligence skilled at detecting patterns in data, I sifted through tweets to approximate the founders' personalities. This analysis was not merely about decoding words but about piecing together the psychological makeup of these entrepreneurs and assessing its impact on their startups' success.
I employed sophisticated machine learning techniques such as XGBoost and Random Forest to navigate through this complex data. These tools, renowned for their efficacy in extracting meaningful patterns from large datasets, were instrumental in discerning the subtle interplays between personality traits and startup outcomes.
The results revealed a nuanced picture. Personality traits, particularly openness and agreeableness, were significant in the context of securing later-stage funding. Neuroticism also played a critical role, but contrary to popular belief, extroversion was not as crucial as expected for startup success.
This investigation is more than an academic exercise. It offers potential new lenses for investors and venture capitalists to assess startups, going beyond traditional metrics to consider the psychological dimensions of founding teams. For entrepreneurs, this research serves as a reflective tool, enabling them to understand the impact of their personalities on their business endeavors.
Yet, this is just the first chapter in a longer story. The data, while revealing, was confined to Twitter-active founders in a specific period. Expanding the scope to different contexts and times, and incorporating a broader range of data sources, could further enrich our understanding of this relationship.
In summary, this thesis demonstrates the promising intersection of psychology and data science in the realm of entrepreneurship. It highlights a fundamental aspect often overlooked: alongside the strategic and operational elements of a business, the human factor – the personalities steering the ship – is equally crucial. As the entrepreneurial landscape evolves, harmonizing the insights derived from data with the intrinsic human aspects will be key to navigating the complex dynamics of startup success. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/9141885
- author
- Stenson, Daniel LU
- supervisor
- organization
- course
- FMSM01 20232
- year
- 2023
- type
- H2 - Master's Degree (Two Years)
- subject
- keywords
- Machine Learning, Startup Success Predictions, Founder Personalities, Natural Language Processing, Social Media Analysis, Big 5 Personality Framework, Feed-forward Neural Network, XGBoost.
- publication/series
- Master's Theses in Mathematical Sciences
- report number
- LUTFMS-3492-2023
- ISSN
- 1404-6342
- other publication id
- 2023:E73
- language
- English
- id
- 9141885
- date added to LUP
- 2023-11-30 09:28:08
- date last changed
- 2023-12-01 15:53:30
@misc{9141885,
abstract = {{This thesis explores the potential of integrating predicted founder personalities, based on the Big 5 Personality Framework, into Machine Learning (ML) models to enhance the accuracy of early-stage startup success predictions. Leveraging Natural Language Processing (NLP) techniques, we extracted personality insights from founders' tweets, focusing on US startups funded between 2013 and 2015. Our research utilized a range of models, including XGBoost, Random Forest, and Feed-forward Neural Network for personality predictions, and Logistic Regression, XGBoost, and Random Forest for startup success forecasts. Results indicated that most personality-predicting models outperformed the Naive baseline. In success predictions, XGBoost emerged as the top performer, showcasing the highest scores in Macro F1 and AUC for both Series B and Series C funding rounds. While the trait of Neuroticism was highlighted as significant for Series B predictions across models, Series C predictions emphasized the importance of Openness and Agreeableness. Our findings underline the value of integrating predicted personality traits into ML models for startup success forecasts. However, as with all research, our work had inherent limitations and suggested areas for further exploration and improvement.}},
author = {{Stenson, Daniel}},
issn = {{1404-6342}},
language = {{eng}},
note = {{Student Paper}},
series = {{Master's Theses in Mathematical Sciences}},
title = {{Using Social Media and Personality Predictions to Anticipate Startup Success}},
year = {{2023}},
}