17.11.2026
Prof. Dr Ganna Demydyuk receives US research award
AI-powered analysis of customer reviews aims to help businesses
How can companies analyse thousands of customer reviews to identify why customers stay, recommend the business or switch to a competitor? Dr Ganna Demydyuk, a professor based in Bremerhaven, and Dr Waqas Ali from the University of Wolverhampton have been investigating this question. Using artificial intelligence, they developed an approach that translates unstructured customer comments into structured behavioural data, thereby making it usable for management and controlling systems. For this research, they were recently awarded the Best Paper Award at the Global Conference on Technology and AI in Hospitality and Tourism (GLOTECH).
Whether in the form of a Google review or on platforms such as Trustpilot, customer reviews can influence whether other people go on to buy certain products or use specific services. For this reason, they are important sources of data for businesses. They help to understand what consumers’ needs are and what motivates them to make a purchase. For Professor Dr Ganna Demydyuk from Bremerhaven, this data is also fascinating from an academic perspective. She teaches on the International Tourism Management (ITM) programme and, as a researcher, aims to help businesses achieve economic success through her analyses. “Companies today have access to enormous amounts of customer feedback. The real challenge lies not in collecting this data, but in analysing it in such a way that it yields actionable insights for management decisions. Our approach uses artificial intelligence to translate customer comments into structured behavioural signals that can be linked to operational and financial metrics,” explains Prof. Demydyuk.
She says that her use of artificial intelligence is what makes her work special: whilst AI is increasingly being used in behavioural research, its potential for addressing business management issues has so far been relatively under-exploited. The innovative aspect of the research lies not in the use of AI alone, but in how it is applied. “We use AI to systematically derive information about loyalty, churn risks and future behaviour from unstructured customer narratives. This means that, for the first time, customer reviews can be utilised as part of a continuous management and control system,” says the professor.
Good service prevents churn; emotional experiences foster loyalty
In their latest research, the researchers investigated how companies can glean insights into customer loyalty, churn risks and future demand from customer reviews. The case study drew on 183,991 English-language cruise reviews from the cruiseline.com platform, covering the years 2002 to 2023. However, the aim of the study was not to analyse a single sector, but to develop a transferable model for the AI-assisted evaluation of customer experiences.
Using AI-assisted text analysis and statistical models, the researchers investigated how different experience characteristics correlate with mood, satisfaction and expressed behavioural intentions. “Cruises are particularly well suited to such analyses because customers provide very detailed and emotional accounts of their experiences. However, the methods developed here can, in principle, be applied to many other sectors, such as the hotel industry, the hospitality sector, retail, transport or e-commerce,” explains Prof. Demydyuk. A key finding of the study is that satisfaction alone is not a reliable indicator of customer loyalty. What is far more crucial is how customers perceive the overall experience, process it emotionally and assess it in terms of value for money. “The results show that customer reviews should be analysed not only in terms of factual content, but also in terms of their emotional content and the behavioural cues they contain,” says the professor. The analysis makes it clear that travellers often discuss aspects other than those that actually influence their loyalty or their intention to book again. Fundamental service characteristics such as reliability, organisation or service quality primarily prevent customers from switching to competitors. Long-term customer loyalty, on the other hand, often arises from emotionally charged experiences, such as special moments, social interactions, cultural impressions or the feeling of having experienced something unique. “This is particularly relevant for businesses because customer comments thus provide far more than just satisfaction scores. They contain clues to future customer behaviour, which can be systematically identified with the help of AI and utilised to inform management decisions.”
Model designed to support businesses
Based on their analyses, the researchers developed a multi-layered model that links customer comments, emotional reactions, satisfaction ratings and behavioural intentions. The aim is to convert unstructured text data into decision-relevant information that can be integrated with operational metrics, management reports and controlling systems. Customer reviews contain not only opinions but also leading indicators of future behaviour. The challenge lies in making these signals visible and utilising them for management decisions. However, the actual economic impact of customer reviews still needs to be verified: the study does not contain any turnover or repeat booking data and therefore does not demonstrate either higher revenue or the successful prevention of customer churn. In the future, it would be possible to link the feedback with internal transaction and accounting data and analyse changes over time. The results of the study have already impressed the jury of the Best Paper Award at the Global Conference on Technology and AI in Hospitality and Tourism. Ganna Demydyuk and Waqas Ali were delighted to receive this year’s award for their paper ‘A Multi-Layer Model of Customer Experience Management for Continuous Data Auditing’.
The study was conducted as part of the cruise research programme at Bremerhaven University of Applied Sciences, whose long-standing expertise in the fields of cruise management, tourism and the maritime economy was a key factor in Prof. Dr Ganna Demydyuk’s decision to join the university. Participation in the conference and the presentation of the research findings were supported by the university’s research fund.