A machine learning analysis of 24/7 coffee shop reviews: evidence from Ho Chi Minh city, Vietnam
The‐Bao Luong
What the paper says
Purpose This study examines customer sentiment and experiential dynamics in 24/7 coffee shops in Ho Chi Minh City, Vietnam, using large-scale user-generated content from Google Maps reviews. Design/methodology/approach A mixed-methods framework integrates natural language processing, sentiment analysis, topic modeling, and machine learning. The analysis draws on a single-platform dataset of 35,366 Google Maps reviews from 24/7 coffee shops in Ho Chi Minh City to identify sentiment patterns, thematic structures, and predictive capabilities. While Google Maps was selected for its extensive coverage and high review volume, reliance on a single source is acknowledged as a limitation and discussed in the concluding section. Findings The results reveal that 72.1% of reviews express positive sentiment, highlighting themes of friendly staff, a comfortable ambiance, and high-quality beverages. Negative reviews, comprising 16.7% of the dataset, highlight recurring issues, including poor service attitudes, environmental discomfort, and perceived overpricing. Sentiment varies by time: positive peaks midday, whereas negative rises after midnight, reflecting operational challenges during late hours. Topic modeling and manual thematic analysis identified fourteen distinct themes across sentiment categories, reflecting both emotional and functional dimensions of the customer experience. A Random Forest classifier achieved 94.91% accuracy in predicting sentiment, demonstrating the effectiveness of machine learning in automating review analysis. Practical implications Operators can apply these insights by improving staff training for late-night service, upgrading infrastructure such as Wi-Fi and air conditioning, and leveraging AI-driven sentiment monitoring for real-time feedback management. Originality/value This research advances experiential consumption and urban coffee culture theory by extending customer experience frameworks into continuous service environments. Unlike prior studies focused on daytime café culture or single-method approaches, it combines computational and qualitative techniques to uncover nuanced experiential patterns. It introduces a scalable, data-driven model for theorizing urban hospitality in emerging Asian megacities.
Evidence weight
Balanced mode · F 0.40 / M 0.15 / V 0.05 / R 0.40
| F · citation impact | 0.50 × 0.4 = 0.20 |
| M · momentum | 0.50 × 0.15 = 0.07 |
| V · venue signal | 0.50 × 0.05 = 0.03 |
| R · text relevance † | 0.50 × 0.4 = 0.20 |
† Text relevance is estimated at 0.50 on the detail page — for your query’s actual relevance score, open this paper from a search result.