SEGMENTATION OF TRAC-ASTRA RENT A CAR CUSTOMERS BASED ON HISTORICAL BEHAVIOR AS A BASIS FOR DETERMINING PROMOTION

• Main Author: Daniasya, Widiyana
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/79267
• Institution: Institut Teknologi Bandung

Customers are important assets for every business. Therefore, companies need to retain as many existing customers as possible. TRAC – Astra Rent A Car is a car rental company with one of the services offered is short-term rental for individual customers. TRAC currently utilizes promotion programs as a part of customer retention efforts for individual customers, yet current promotions are considered ineffective due to the lack of knowledge about individual customers’ characteristics, one of which is behavioral characteristics. This research employs latent class analysis method to form customer segments based on customers’ actual historical behavior in renting cars. The variables used as segmentation bases consist of rental frequency, reservation lead time, rental in high season period, proportion of weekend rental, rental duration, and vehicle type. The data used for this research is secondary data in the form of rental transaction data ranging from January 2019 to May 2021. The aforementioned data is obtained from the company’s database. The segmentation process using latent class analysis results in five customer segments, namely ‘infrequent business traveler’, ‘frequent customer’, ‘short-term weekend leisure traveler’, ‘one-time holiday customer’, and ‘go-show business traveler’. Promotions are proposed for two chosen segments, which consist of ‘frequent customer’ and ‘one-time holiday customer’. The proposed promotion for ‘frequent customer’ is patronage reward, while the proposed promotion for ‘onetime holiday customer’ is special holiday offer with complimentary voucher.