Identifying high-value keywords for Bali tourism: A machine learning approach

This study explores how machine learning techniques, specifically K-Means Clustering, can be applied to identify high-value tourism-related keywords for Bali using Google Keyword Planner data. By analyzing normalized search volume, cost-per-click (CPC), and competition scores, the keywords were grouped into three meaningful clusters: (1) high-volume, high-CPC keywords such as Bali luxury resorts and honeymoon in Bali, (2) low-volume, high-CPC keywords like eco-retreat Bali and private villa Ubud, and (3) high-volume, low-CPC keywords including Bali itinerary and Bali beach names. The Elbow Method confirmed that three clusters offered the optimal segmentation. Findings show that luxury and culturally immersive travel themes yield the highest marketing value, while niche keywords offer untapped advertising potential. Seasonal analysis revealed peaks in keyword interest during mid-year and year-end holiday seasons, aligning with international travel patterns. The machine learning approach enhanced keyword structuring and revealed strategic timing opportunities for content deployment. The study concludes that data-driven keyword targeting can significantly improve the effectiveness and sustainability of digital marketing in Bali’s tourism sector. Recommendations include focusing on premium and niche clusters, while future research may integrate social media trends and geospatial analysis to deepen keyword relevance and behavioral insights.