Natural Language Processing for Corporate Culture Assessment: Lessons Learned for Building a Strong Employee Value Proposition with GPTW Switzerland AG

Demographic change is exacerbating the shortage of skilled workers in the global labor market. Companies have difficulties recruiting and retaining employees in future. Therefore, developing a strong employee value proposition (EVP) and, above all, a good corporate culture is crucial to attract potential employees. However, existing methods for creating and evaluating the EVP are manual and time-consuming. They require a thorough understanding of the company culture and can cost a professional consultant a lot of time to analyze the different cultural elements and values contained in various internal text documents. In this study, we aim to explore the feasibility of using natural language processing (NLP) for partially automated assessment of corporate culture. Together with our business partner GPTW Switzerland AG, we collected responses from over 50 companies to the EVP questionnaire Culture Audit and applied text embedding and topic modeling-based NLP techniques to assess corporate culture. The analytical framework contributes to a sound understanding of how cultural topics, and keywords are used and distributed in companies’ Culture Audit responses. Our findings also highlight the current limitations and potential opportunities for NLP methods to complement the work of consultants in assessing organizational culture, particularly by harnessing the explanatory power of machine learning. Overall, this study could lead to a more synergistic and efficient use of data to provide organizations and their employees with a consistent, transparent, and monitorable employer brand.