Perancangan Strategi Pemasaran Berbasis Analisis Sentimen Naive Bayes

Abstract

Abstract. Aquila Knives is a micro, small, and medium enterprise (MSME) engaged in the production and sale of kitchen knives, butcher knives, outdoor knives, and traditional Indonesian blades. In recent periods, the company experienced a 37% decline in sales, indicating the need to evaluate its marketing strategy based on customer perceptions. This study aims to analyze customer sentiment and develop a marketing strategy based on the analysis results. The dataset consists of 896 YouTube comments collected through web scraping. The analysis process includes text preprocessing, Term Frequency–Inverse Document Frequency (TF-IDF) weighting, sentiment classification using the Naïve Bayes algorithm, and Rule-Based Aspect Extraction based on Kotler’s product quality dimensions: Price, Feature, Reliability, Durability, Performance, Service, and Aesthetic. The results show that the Naïve Bayes model achieved an accuracy of 71.67%. The Performance aspect received the highest positive sentiment, while the Service aspect was dominated by negative sentiment, making it the highest priority for improvement. Based on these findings, a marketing strategy was developed using the Marketing Mix (4Ps) and translated into the 7 StoryBrand framework to support content marketing. The proposed approach provides a more targeted, customer-oriented marketing strategy with the potential to improve marketing effectiveness and increase Aquila Knives' sales.

Abstrak. Aquila Knives merupakan usaha mikro, kecil, dan menengah (UMKM) yang bergerak di bidang produksi dan penjualan pisau, seperti pisau dapur, pisau jagal, pisau outdoor, dan senjata tradisional Indonesia. Dalam beberapa periode terakhir, perusahaan mengalami penurunan penjualan sebesar 37%, yang mengindikasikan perlunya evaluasi strategi pemasaran berdasarkan kebutuhan pelanggan. Penelitian ini bertujuan menganalisis sentimen pelanggan serta merancang strategi pemasaran berbasis hasil analisis tersebut. Data penelitian berupa 894 komentar YouTube yang diperoleh melalui web scraping. Tahapan analisis meliputi preprocessing, pembobotan kata menggunakan TF-IDF, klasifikasi sentimen dengan algoritma Naïve Bayes, serta Rule-Based Aspect Extraction berdasarkan aspek kualitas produk Kotler, yaitu Price, Feature, Reliability, Durability, Performance, Service, dan Aesthetic. Hasil penelitian menunjukkan bahwa Naïve Bayes menghasilkan akurasi sebesar 71,67%. Aspek Performance memperoleh sentimen positif tertinggi, sedangkan aspek Service didominasi sentimen negatif sehingga menjadi prioritas perbaikan. Berdasarkan hasil tersebut, disusun strategi pemasaran menggunakan Marketing Mix (4P) yang diterjemahkan ke dalam kerangka 7 Story Brand sebagai dasar content marketing. Hasil penelitian menunjukkan bahwa pendekatan ini menghasilkan strategi pemasaran yang lebih terarah, berbasis kebutuhan pelanggan, dan berpotensi meningkatkan efektivitas pemasaran serta penjualan Aquila Knives.