Method and system for building a consumer decision tree in a hierarchical decision tree structure based on in-store behavior analysis
View Patent ↗The present invention is a system and method for determining the hierarchical purchase decision process of consumers in front of a product category. The decision path of consumers is obtained by combining behavior data with the category layout and transaction data based on observed actual in-store purchase behavior using a set of video cameras and software for extracting sequence and timing of each consumer's decision process. A hierarchical decision tree structure comprises nodes and edges, wherein a node represents the state-of-mind of the consumer, the number of nodes is predefined, and an edge represents the transition of the decision. The decisions for each product group are captured down to the product attribute level and analyzed by demographic group. The outcome provides relative importance of each product attribute in the purchase decision process, and helps retailers and manufacturers to evaluate the layout of the category and customize it for key segment.
1. A method for building a consumer decision tree based on in-store purchase behavior analysis by the measurement of a set of consumer behavior metrics,
comprising the following steps of:
a) capturing a plurality of input images of consumers by at least a means for capturing images in a store area,
b) processing the plurality of input images in order to analyze the behavior of the consumers,
c) measuring decision activities of the consumers tied to product categories based on the behavior analysis,
d) creating a plurality of datasets by accumulating the decision activities, whereby decision activity is measured based on the actual in-store purchase behavior of the consumers including interaction with products and travel paths to categories, as opposed to using intercepts or panels to develop them, and
e) constructing a hierarchical decision tree structure, clustering the consumer behavior data based on the measurement of the decision activities by the consumers, which comprises nodes and edges,
wherein a node represents in-store purchase behavior of the consumer,
wherein the number of nodes is predefined, and
wherein an edge represents the transition of the decision activities.
2. The method according to claim 1 , wherein the method further comprises a step of objectively capturing a large amount of statistical data, whereby the gathered data is unbiased.
3. The method according to claim 1 , wherein the method further comprises a step of calculating which decision was made by the majority of consumers based on a large number of gathered samples.
4. The method according to claim 1 , wherein the method further comprises a step of calculating a dominant path among decision paths in the consumer decision tree.
5. The method according to claim 1 , wherein the method further comprises a step of constructing a decision sub-tree for each consumer.
6. The method according to claim 1 , wherein the method further comprises a step of organizing decision paths for consumer segments by demographic class or by need states,
wherein attributes for the demographic class comprise age, gender, and ethnicity.
7. The method according to claim 1 , wherein the method further comprises a step of counting selection activities according to the category, product, brand, size of product package made by the decision activities, based on spatial and temporal observation using video analytics.
8. The method according to claim 1 , wherein the method further comprises steps of:
a) organizing products in a coordinate system according to predefined attributes,
b) measuring what attributes in the attribute block are chosen, and
c) measuring the sequence of the chosen attributes,
wherein the attributes comprise brand, product type, size, flavor, and price.
9. The method according to claim 1 , wherein the method further comprises a step of developing a planogram that matches the decision process of the consumer as well as their needs.
10. The method according to claim 1 , wherein the method further comprises a step of managing categories based on the consumer decision tree,
wherein category management comprises sku rationalization, assortment of categories, and efficient space allocation.
11. The method according to claim 1 , wherein the method further comprises a step of developing product assortments based on an understanding of a hierarchy of category decisions.
12. The method according to claim 1 , wherein the method further comprises a step of processing allocation of retail space based on the consumer decision tree.
13. The method according to claim 1 , wherein the method further comprises a step of integrating a semi-automated method to compute the consumer decision tree to identify behaviors that cannot be automatically observed or to identify granular details of consumer interaction with products.
14. The method according to claim 1 , wherein the method further comprises a step of superimposing price elasticity on behavior, based on point-of-sale data from several price points.
15. The method according to claim 1 , wherein the method further comprises a step of detecting a sequence of decision activities that are made by a consumer,
wherein the sequence of decision activities is a decision path from a node to the next node, and
wherein the sequence of decision activities shows end-to-end decisions made by the consumer.
16. The method according to claim 1 , wherein the method further comprises a step of optimizing marketing based on the consumer decision tree.
17. The method according to claim 1 , wherein the method further comprises a step of measuring different levels of decision steps based on lengths of decision paths,
wherein average time of making a decision can be calculated based on the levels of decision steps.
18. The method according to claim 1 , wherein the method further comprises a step of measuring a premeditated decision compared to impulse decision based on the consumer decision tree.
19. The method according to claim 1 , wherein the method further comprises a step of quantifying a percentage of consumers for each node.
20. An apparatus for building a consumer decision tree based on in-store purchase behavior analysis by the measurement of a set of consumer behavior metrics, comprising:
a) at least a means for capturing images that captures a plurality of input images of consumers in a store area,
b) at least a computer programmed to execute the following steps of:
processing the plurality of input images in order to analyze the behavior of the consumers,
measuring decision activities of the consumers tied to product categories based on the behavior analysis,
creating a plurality of datasets by accumulating the decision activities,
whereby decision activity is measured based on the actual in-store purchase behavior of the consumers including interaction with products and travel paths to categories, as opposed to using intercepts or panels to develop them, and
constructing a hierarchical decision tree structure, clustering the consumer behavior data based on the measurement of the decision activities by the consumers, which comprises nodes and edges,
wherein a node represents in-store purchase behavior of the consumer,
wherein the number of nodes is predefined, and
wherein an edge represents the transition of the decision activities.