System and method for fine-tuning sales clusters for stores

Patent metadata

Field	Value
Publication number	US-11276033-B2
Application number	US-201816205604-A
Country	US
Kind code	B2
Filing date	Nov 30, 2018
Priority date	Dec 28, 2017
Publication date	Mar 15, 2022
Grant date	Mar 15, 2022

Abstract

Based upon the transition information for all the retail stores, a determination is made of an average cluster retention score. The average cluster retention score is a measure of how many retail stores have moved from original to different existing sales clusters in the current sales period. When the average cluster retention score is below a predetermined threshold, a complete re-organization of the existing sales clusters is performed. When the average cluster retention score is above the predetermined threshold, a determination is made as to whether each retail store should be re-classified.

First claim

What is claimed is:

1. A system for updating planograms that are used to arrange products in retail stores, the system comprising:
a plurality of sales entry devices that are configured to collect current sales data at a plurality of retail stores during a current sales period;
a transceiver circuit disposed at a central processing center, the transceiver circuit being configured to receive the current sales data;
a database disposed at the central processing center that is configured to store the current sales data and additionally store a representation of existing sales clusters of retail stores, each of the existing sales clusters including a subset of the plurality of retail stores such that each retail store belongs to one of the existing sales clusters, each of the existing sales clusters being assigned an existing planogram;
a control circuit that is disposed at the central processing center, the control circuit being coupled to the transceiver circuit and the database, the control circuit being configured to:
based upon the current sales data, determine transition information that identifies retail stores that have potentially moved from an original one of the existing sales clusters to a different one of the existing sales clusters, wherein the transition information relates to Euclidian distances between retail stores that have moved and the existing sales clusters;
based upon the transition information for all the retail stores, determine an average cluster retention score, the average cluster retention score being a numeric measure of how many retail stores have moved from original to different existing sales clusters in the current sales period;
when the average cluster retention score is below a predetermined threshold, perform a complete re-organization of the existing sales clusters based at least in part upon the current sales data, wherein the complete reorganization does not necessarily keep any of the existing sales clusters;
when the average cluster retention score is above the predetermined threshold, determine whether each retail store should be re-classified as belonging to a different existing sales cluster or should be placed in a newly created sales cluster;
determine an optimal planogram for each retail store based at least in part upon whether a complete re-organization is performed, whether the retail store is re-classified as belonging to a different existing sales cluster, or whether the retail store is placed in a newly created sales cluster, the optimal planogram when implemented being effective to maximize store performance;
selectively implement an action based upon the optimal planogram, the action at least in part using an automated vehicle or robot to stock shelving units in the stores according to the planogram, the stocking being accomplished by moving or rearranging products within the retail stores to the shelving units in the retail stores, the action also including reconfiguring at least one of the shelving units in at least one of the plurality of retail stores;
wherein the determination of whether retail stores are re-classified or placed in a newly created cluster is based upon a confidence score, the confidence score indicating the likelihood a particular retail store belongs to an existing sales cluster and wherein the confidence score is indicative of a distance of the particular store to the existing sales cluster;
wherein a first value of the confidence score indicates that the particular retail store is closer to the existing cluster compared to the other clusters, wherein a second value of the confidence score that is less that the first value indicates the particular retail store is at a similar distance as between two clusters, and wherein a third value of the confidence score that is lower than the second value indicates that particular retail store should be place in the newly created cluster.

2. The system of claim 1 , wherein the transition information is organized according to a transition matrix that indicates transitions of the retail stores between different ones of the existing sales clusters.

3. The system of claim 1 , wherein the optimal planogram is an existing planogram or a newly-created planogram.

4. The system of claim 1 , wherein the existing sales clusters each have a cluster center, and the retail store is re-classified to the existing sales cluster having the closest cluster center to the retail store according to the current sales data.

5. The system of claim 1 , wherein the current sales data is this-year sales data.

6. A method for updating planograms that are used to arrange products in retail stores, the method comprising:
collecting current sales data at a plurality of sales entry devices at the plurality of retail stores during a current sales period;
receiving the current sales data at a central processing center;
storing the current sales data and a representation of existing sales clusters of retail stores at a database disposed at the central processing center, each of the existing sales clusters including a subset of a plurality of retail stores such that each retail store belongs to one of the existing sales clusters, each of the existing sales clusters being assigned an existing planogram;
based upon the current sales data and at a control circuit that is disposed at the central processing center, determining transition information that identifies retail stores that have potentially moved from an original one of the existing sales clusters to a different one of the existing sales clusters, wherein the transition information relates to Euclidian distances between retail stores that have moved and the existing sales clusters, based upon the transition information for all the retail stores, determining at the control circuit an average cluster retention score, the average cluster retention score being a numeric measure of how many retail stores have moved from original to different existing sales clusters in the current sales period, when the average cluster retention score is below a predetermined threshold, performing at the control circuit a complete re-organization of the existing sales clusters based at least in part upon the current sales data, wherein the complete reorganization does not necessarily keep any of the existing sales clusters, when the average cluster retention score is above the predetermined threshold, determining at the control circuit whether each retail store should be re-classified as belonging to a different existing sales cluster or should be placed in a newly created sales cluster, determining at the control circuit an optimal planogram for each retail store based at least in part upon whether a complete re-organization is performed, whether the retail store is re-classified as belonging to a different existing sales cluster, or whether the retail store is placed in a newly created sales cluster, the optimal planogram when implemented being effective to maximize store performance, and at the control circuit selectively implementing an action based upon the optimal planogram, the action at least in part using an automated vehicle or robot to stock shelving units of the stores according to the planogram, the stocking being accomplished by moving or rearranging products within the retail stores to the shelving units in the retail stores, the action also including reconfiguring at least one of the shelving units in at least one of the plurality of retail stores;
wherein the determination of whether retail stores are re-classified or placed in a newly created cluster is based upon a confidence score, the confidence score indicating the likelihood a particular retail store belongs to an existing sales cluster and wherein the confidence score is indicative of a distance of the particular st

Assignees

• Walmart Apollo Llc

Inventors

• Karmakar Somedip
• Manna Sourit
• Pal Gayatri

Classifications

• G06Q30/0204

  Market segmentation · CPC title

• G06Q10/087 Primary

  Inventory or stock management, e.g. order filling, procurement or balancing against orders · CPC title