A Visual Tool to Understand and Predict Consumers' Expectations Through Ideal Point Modeling

Transcription

1 2012 Annual Meeting of Korean Society of Food Science and Technology June 2012 Convention Center, Daejeon, South Korea A Visual Tool to Understand and Predict Consumers' Expectations Through Ideal Point Modeling Benoît Rousseau, Ph.D. The Institute for Perception Richmond, VA, USA Benoit.Rousseau@IFPress.com /33 Where are we going? Contents Liking and ideal point modeling Landscape Segmentation Analysis Background Application examples Properties of soap bar images Adults and children food preferences Conclusions 2/33

2 2012 Annual Meeting of Korean Society of Food Science and Technology June 2012 Convention Center, Daejeon, South Korea A Visual Tool to Understand and Predict Consumers' Expectations Through Ideal Point Modeling What is liking? 3/33 Drivers of Liking : The Problem Product developers need actionable guidance to create and improve products Consumers are not calibrated to provide consistent descriptive information Chocolate 4 Caramel 8 Caramel 6 Caramel 8 Fruity 1 Vanilla 2 Caramel 7 Caramel 7 Fresh 5 Sweet 10 Caramel 7 Caramel 7 Caramel 8 Almond 10 Caramel 7 Caramel 7 Consumer panel Trained panel 4/33

3 Two Different Modeling Approaches : 212 LSA LSA IPM : 26 IPM Increasing average liking N P Y D M R U O W B A K V X E H C S Q J T L F G I Z 5/33 What is Liking? Liking is not a sensory variable Bitter Salty Sweet Liking 6/33

4 Concept of Ideal Point Coombs proposed a liking model based on distance from ideal Liking Ideal Point Ideal Point Sweetness 7/33 Landscape 2012 Annual Meeting of Korean Society of Food Science and Technology June 2012 Convention Center, Daejeon, South Korea A Visual Tool to Understand and Predict Consumers' Expectations Through Ideal Point Modeling Segmentation Analysis 8/33

5 Step 1-1 Map Generation LSA first unfolds liking and creates a space relevant to consumer acceptability (6 products, 44 consumers) The closer a consumer is to a product, the more he/she likes it Contours indicate consumer densities and facilitate the visualization of potential segmentation Consumers P 1 P 4 P 2 P 5 P 1 P 6 P 3 P 4 P 3 P 6 P 2 P 5 9/33 Step 2 - Finding the Drivers of Liking Example: P 4 P 6 P 3 P 1 P 5 P 2 Expert Sweet attribute P 1 P 4 P 2 P 5 P 1 P 3 P 6 Sweet P 2 P 4 P 3 P 6 P 5 10/33

6 Step 2 - Finding the Drivers of Liking (cont.) Some attributes can be fit on the map and are drivers of liking Others can t and are less relevant to consumer acceptability Crunchy P 1 P 4 Berry P 1 P 6 P 2 P 5 P 3 P 4 Sweet P 2 P 3 P 6 P 5 Smooth Vanilla 11/33 Step 3 - Finding Optimal Products Based on the ideal point distribution, optimal product locations can be estimated Using the drivers of liking, each of the optimal profiles can then be generated, providing valuable guidance for subsequent product development Crunchy Optimal profiles on 7-point intensity scales Berry Sweet Smooth Vanilla Smooth Crunchy Vanilla Berry Sweet /33

7 Step 4 - Predicting New Product Performance Using new products profiles, their approximate locations on the LSA map can be estimated Crunchy Products A and B profiles on 7-point intensity scales Product A Product B Berry Smooth 6.3 A Sweet Crunchy Sweet Smooth B Vanilla Berry Vanilla A B 13/ Annual Meeting of Korean Society of Food Science and Technology June 2012 Convention Center, Daejeon, South Korea A Visual Tool to Understand and Predict Consumers' Expectations Through Ideal Point Modeling Landscape Segmentation Analysis Applications 14/33

8 A Visual Tool to Understand and Predict Consumers' Expectations Through Ideal Point Modeling Expected Moisturizing and Refreshing Properties of Soap Bar Images 15/33 Study Design Pictures of bar soaps rated on moisturizing and refreshing expectation 25 different pictures, central composite design on 4 variables (Translucency, Shine, Hue and Saturation), 5 levels per variable Total of 31 pictures (middle point evaluated 7 times) Translucency Translucency Hue Saturation Shine Shine 16/33

9 31 Soap Images 7x 17/33 Study Design (continued) Study conducted in the Tokyo area in Japan 610 female consumers, split in two groups Group 1: Moisturizing (310 consumers) Group 2: Refreshing (300 consumers) Expected Moisturizing and Refreshing properties rated on a 15-point scale Sessions lasted about 30 minutes 18/33

10 Study Design (continued) Respondent Interviewer 19/33 Moisturizing LSA Soap image Consumer Individual points show the location of the ideal moisturizing soap image for each consumer x /33

11 Moisturizing vs. Refreshing LSAs Moisturizing LSA Refreshing LSA 21/33 Expected Moisturizing and Refreshing Properties of Soap Bar Images Drivers of Perception 22/33

12 Perception: Drivers Translucency Shine Hue Saturation Moisturizing Refreshing Translucency Saturation 23/33 Expected Moisturizing and Refreshing Properties of Soap Bar Images Optimal Products 24/33

13 Moisturizing Perception Optimum Optimum Moisturizing 7x 25/33 Refreshing Perception Optimum Optimum Refreshing 26/33

14 Expected Moisturizing and Refreshing Properties of Soap Bar Images Creating the Optimal Products 27/33 Moisturizing Perception: Optimum Optimum Moisturizing Translucency x Saturation 28/33

15 Refreshing Perception: Optimum Translucency Saturation 29/33 Children and Adults Food Preferences 30/33

16 Children and Adult Food Preferences Liking for 19 foods by 150 adults and 150 children (8-12 years old) Only names given, no actual tasting of the foods Apple sauce Chocolate milk Fruits Orange juice Soda Bottled water Cookies Hamburger Pizza Soup Cup cakes Ice cream Popsicle Spaghetti Chicken French fries Iced tea Sandwich Tossed salad Adults liking for foods for their children 31/33 Children and Adult Food Preferences Children Adults Hamburger Chicken Spaghetti Soup Iced tea French fries Tossed salad Pizza Cup cakes Soda Sandwich Cookies Ice cream Fruits Popsicle Chocolate milk Bottled water Orange juice Apple sauce 32/33

17 2012 Annual Meeting of Korean Society of Food Science and Technology June 2012 Convention Center, Daejeon, South Korea A Visual Tool to Understand and Predict Consumers' Expectations Through Ideal Point Modeling Thank You Very Much Any Questions? Benoît Rousseau, Ph.D. The Institute for Perception Richmond, VA, USA Benoit.Rousseau@IFPress.com /33