Presentation Information
[5Yin-A-03]Time-Series Analysis of Customer Demographics of Restaurant Chains Based on Mobile Carrier Search Logs
〇Daichi Inoue1, Hayato Mori2, Taketoshi Yoshii1, Sadakazu Kinoshita1, Daichi Mochihashi3 (1. D2C Inc, 2. NTT DOCOMO, INC, 3. The Institute of Statistical Mathematics)
Keywords:
Search Logs,Normalized Pointwise Mutual Information,Time-Series Analysis,Restaurant Chains,Customer Demographic Analysis
Although search logs reflect voluntary user interests, specific target demographic data for brands is rarely public. This study analyzes temporal interest demographics for restaurant chains using search logs and basic gender/age data from a major Japanese telecommunications carrier. By calculating Normalized Pointwise Mutual Information (NPMI) for co-occurring query terms, we quantitatively identified interest across different user segments. Furthermore, visualizing monthly NPMI trends revealed shifts in interest structures driven by limited-time offers, collaborations, and social events. Results indicated that while broad customer bases exist per restaurant category, distinct demographic variations and temporal fluctuations occur among specific chains within the same category. This method allows for tracking demographic changes without relying on visitation or purchase data, demonstrating significant potential for analyzing marketing effectiveness and competitive landscapes.