Cross-Selling Strategies in e-Marketing 1. Market Basket Analysis Correlation Matrix Factor Analysis 2. Profiling and Cloning Customers Automatic Cluster Detection 3. Targeting the Best Customers The Linkage between Product and Customer Classification – Automatic Interaction Detection 2019/4/29
決策模式庫 Example Rules for Selecting Marketing and Advertising Objectives Marketing Objectives IF product life cycle stage = introduction AND innovation type = discontinuous THEN marketing objective = stimulate primary demand IF brand usage = none THEN marketing objective = simulate brand trial IF current brand usage = some AND (brand switching = high OR product usage rate = fixed) THEN marketing objective =simulate repeat purchase/loyalty IF current brand usage = some AND brand switching =low AND product usage rate = variable THEN marketing objective = increase rate of brand usage 2019/4/29
Advertising Objectives IF marketing objective = simulate primary demand AND product purchase motivation direction = negative THEN ad objective = convey product category information IF marketing objective = simulate brand trial AND brand purchase motivation direction = positive THEN ad objective = convey brand image IF time of brand decision = at point of purchase AND package visibility = high AND package recognition = low THEN ad objective = increase brand recognition IF marketing objective = increase rate of brand usage AND new brand use = yes THEN ad objective = convey new brand uses 2019/4/29
各類書籍之購買程度矩陣 2019/4/29
各類書籍之購買程度矩陣 2019/4/29
各類書籍之購買程度矩陣 2019/4/29
決策模式庫 IF 性別＝女 AND 年齡＝{31,40} AND 教育程度＝專科 THEN 推薦書籍＝{室內設計，攝影} 2019/4/29
自動互動檢視法（AID）理論模式 自動互動檢視法（Automatic Interaction Detector）是Sonquist及Morgan於1963年所提出一種歸納性的（Inductive）的多變量統計分析方法。AID分析之基本原理是反覆的運用單一因子變異數分析（One-way ANOVA）之技術，以尋求一組預測變數及其次分類項（Categories）的組合，以此組合將原始樣本分割（Split）成若干互斥的次群體，使得這些群體之準則變數的平均值差異最大，而達到差異極大化之區隔目的。並且經由以樹狀圖表示的AID分析結果，研究者可以判定預測變數是否有互動關係存在，而有助於研究者設立準則變數與預測變數間最適的函數關係。 2019/4/29
2019/4/29
2019/4/29
2019/4/29
Automatic Interaction Detection B:籃球,網球 C:慢跑,羽球 D:汰舊換新,過年 E:開學,逛街非計劃 因新廣告的影響 F:1200元以下 G:1201元以上 H:800元以下 I:801元以上 J:25~45歲 K:24歲以下,46歲 以上 E 購買預算 B 購買時機 H D 最常做的運動 A G C 購買預算 K F 年齡 J 2019/4/29