MEMBACA NIAT KUNJUNGAN KEMBALI MELALUI ANALISIS SENTIMEN PEMASARAN SENSORIK

Abstrak

Persaingan intens di sektor restoran kawasan wisata menuntut diferensiasi beyond kualitas makanan, khususnya melalui pemasaran sensorik yang mempengaruhi keputusan pembelian konsumen di era transformasi digital. Penelitian ini menganalisis sentimen ulasan Google Maps Amarta Penida Restaurant untuk memahami dampak pemasaran sensorik terhadap persepsi dan niat berkunjung kembali konsumen. Metode kuantitatif deskriptif diterapkan melalui analisis sentimen berbasis leksikon VADER pada 1.195 ulasan pelanggan periode 2021 - 2025, analisis TF-IDF untuk identifikasi kata diskriminatif, dan validasi menggunakan algoritma Naive Bayes. Hasil menunjukkan dominasi sentimen positif sebesar 90,0 persen dengan rasio positif-negatif 24:1. Analisis TF-IDF mengidentifikasi kata "view" dan "sunset" sebagai prediktor utama sentimen positif, mengonfirmasi elemen visual merupakan stimulus pemasaran sensorik terpenting. Terdapat kongruensi antara rating numerik tinggi dan sentimen tekstual positif, dengan akurasi validasi Naive Bayes mencapai 91,38 persen. Penelitian mengonfirmasi aplikasi Teori Stimulus-Organism-Response dan Theory of Planned Behavior dalam konteks restoran wisata, memberikan implikasi manajerial untuk optimalisasi pengalaman sensorik dan kontribusi teoretis signifikan bagi literatur pemasaran sensorik.

Intense competition in the restaurant sector of the tourist area demands differentiation beyond food quality, in particular through sensory marketing that influences consumer purchasing decisions in the era of digital transformation. This study analyzed the sentiment of Google Maps reviews of Amarta Penida Restaurant to understand the impact of sensory marketing on the perception and intention of revisiting consumers. Descriptive quantitative methods were applied through sentiment analysis based on VADER lexicon on 1,195 customer reviews for the period 2021-2025, TF-IDF analysis for discriminative word identification, and validation using Naive Bayes algorithm. The results showed a predominance of positive sentiment by 90.0 percent with a positive-negative ratio of 24:1. The TF-IDF analysis identified the words "view" and "sunset" as key predictors of positive sentiment, confirming the visual element is the most important sensory marketing stimulus. There is a congruence between high numerical ratings and positive textual sentiment, with Naive Bayes ' validation accuracy reaching 91.38 percent. The research confirms the application of Stimulus-organism-response theory and Theory of Planned Behavior in the context of restaurant tourism, providing managerial implications for the optimization of sensory experience and significant theoretical contributions to the sensory marketing literature.