ADVISING CASE STUDIES

Using the unobtrusive corporate culture analysis tool (UCCAT), we developed bespoke AI algorithms that analysed the textual comments provided by staff in their engagement survey (> 20 million words). The algorithms tested whether the way employees experience and think in the company was aligned with the aspired corporate values, and we identified localities (e.g., job roles, units) where there was some misalignment, and determined the causes of this. We also measured the overall cultural health of the organisation by determining which norms were most salient in the thoughts and experiences of employees (e.g., for trusting leadership, feeling under pressure to break rules, being achievement-oriented), ascertaining the most positive and negative aspects off culture, and identifying the ‘emerging issues’ (i.e., new topics) and risks that people were experiencing. We qualitatively analysed the text surfaced by the AI-algorithm to develop a psychological profile of the organization, and designed culture interventions targeted at specific units and designed to improve communication and remove barriers to effective job performance.

Based on our Healthcare Complaints Analysis Tool (HCAT) Methodology, we developed a bespoke system for analysing the content and severity of problems reported by service-users to the government department. The system was built through using the HCAT analytical template, and, based on hundreds of complaints sent to the government department, customising it to reflect the specific issues and problems experienced by service users. This meant the government department could not only monitor the number of complaints received, but had a reliable and validated method for tracking and analysing the issues reported by service users over time. The system enabled the government department to identify business areas where the culture for managing service-users required improvement, and to generate concrete insights that could be used to create change. It was integrated into the department’s systems for handling complaints, and was designed so that it could be automated with NLP based measures for measuring the severity and content of negative feedback from service users.

We developed the UCCAT methodology to analyse public data on over 400 companies (e.g., company reports, financial data, board data, online feedback) and develop 108 unobtrusive indicators of culture. An unobtrusive indicator of culture (UIC) refers to a single measure of organizational culture based on naturally occurring data collected without directly engaging employees. Using the UICs, we evaluated the culture of invested companies on dimensions such as employee focus, customer focus, and governance, and identified their comparative strengths and weaknesses to competitor firms (by sector) through assessing them against a published and validated benchmark dataset. This enabled the investment firm to both give feedback to the companies on their culture, and also guide their internal decision-making.

To undertake this project, we surveyed thousands of staff and many organisations (e.g., different airlines) at the airports using our validated safety culture survey. Because air safety relies on all partner organisations to have a strong and shared culture (e.g., aircraft and air traffic control), we examined whether, across the aviation system, organisations had similar and positive norms and values. Our analysis of the survey data, complemented with follow-up interviews and focus groups, provided a fine-grained analysis of each partner’s strengths (e.g., on commitment to safety, risk tolerance), and areas where there was need for improvement within and between organisations (e.g., on communication). Through both the culture measurement process, and engaging with stakeholders, the project led to airports developing a more coherent and positive culture, and better coordination and performance as a result.