尋求多樣化行為(Variety Seeking)和增強性(Reinforcement)忠誠行為對於促銷活動的反 應，以及如何運用尋求多樣化行為和增強性忠誠行為當做市場區隔的變數一直受到許多 行銷學者的重視。過去相關研究的實證資料來源主要是取得消費者個人層級消費序列 (sequence of individual-level purchase data)。但個人層級消費序列資料要夠長才具有統計 意義，但透過實驗法要求消費者回想消費序列資料並不準確，而且很難提供品類(Product Category)及品牌層級(Brand Level)的分析。本研究試圖根據消費者指數研究中之兩個總 體層級的變數(包括: 市佔率、100%忠誠比率)，經由典型馬可夫市佔率模型先估計促銷 效應變數，然後再利用此三個變數(包括: 市佔率、100%忠誠比率及促銷效應變數)導出 一指標用來判別潛在游離消費者是否為尋求多樣化或是增強性忠誠行為。實證資料來源 為TNS（Taylor Nelson Sofres）Global Taiwan 提供之三種快速流動消費性商品(FMCG – Fast Moving Consumer Goods)產品：成人奶粉、洗髮精和洗衣粉之消費者指數研究進行 品類(Product Category)及品牌層級(Brand Level)的分析實證分析。 The research regarding the impact of the behavior of variety seeking and reinforcement on promotional response and marketing variables and the behavioral segmentation based on the behavior of variety seeking and reinforcement had been paid much attention. The major method to obtain empirical data of the measure of variety seeking and reinforcement is to capture the sequences of individual-level purchase data. However, long sequence of individual-level purchase data may not exist because consumers may not have made or recorded many purchases. Further, the data based on the sequences of individual-level purchase can not provide the behavior analysis of category level and brand level data. Therefore, this study attempts to adopt a typical Markov market share model along with a method for estimating the promotional effect on the given market share and loyalty to demonstrate how to develop the index to judge the segment of potential switchers is the behavior of variety seeking or reinforcement. The study employs consumer panel data relating to three categories of goods—adult milk powder, shampoo, and detergent from TNS Global Taiwan in 2008. Finally, we present the analysis of switching behavior of category and brand level.