Retail Demand Transfer Estimation Using Inventory Availability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing demand transfer estimation methods in retail fail to consider out-of-stock conditions, leading to inaccurate inventory management strategies, as they primarily focus on product assortment rather than inventory availability.
Innovation Solution
A processor-implemented method and system that integrates historical sales and inventory data to estimate product demand transfer by determining actual, missing, and expected shares of attribute values, calculating additional shares due to missed values, and computing demand transfer based on these shares, using a multiplicative model to account for inventory availability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing demand transfer estimation methods focus on product assortment, then assortment decisions can be optimized, but out-of-stock conditions are not considered leading to inaccurate inventory management
Solution Approach 1:
The patent segments the demand transfer estimation into two distinct components: assortment-based demand transfer and inventory availability-based demand transfer. By separating these factors, the model can independently analyze and quantify the impact of out-of-stock conditions on demand transfer, thereby improving estimation accuracy while maintaining adaptability to different retail scenarios.
Solution Approach 2:
The patent introduces new parameters to the demand transfer model, specifically incorporating inventory availability metrics (in-stock probability, fill rate) alongside traditional assortment parameters. This parameter expansion allows the model to simultaneously consider both assortment decisions and inventory conditions, resolving the contradiction between estimation accuracy and adaptability to out-of-stock scenarios.
2Measurement precision
If demand transfer estimation includes multiple data sources (sales, inventory, assortment), then estimation accuracy improves, but system complexity increases
Solution Approach 1:
The patent develops a unified demand transfer estimation model that serves multiple functions: it can estimate demand transfer due to assortment decisions, demand transfer due to out-of-stock conditions, and overall demand transfer. This multi-functional approach integrates multiple data sources (sales data, inventory data, assortment data) into a single coherent framework, improving measurement precision while managing system complexity through consolidation rather than separate analysis systems.
Solution Approach 2:
The patent introduces a probabilistic framework as an intermediary mechanism that connects multiple data sources to the demand transfer estimation. This probabilistic model acts as a mediator that systematically processes sales data, inventory data, and assortment data, transforming them into coherent demand transfer estimates. The intermediary framework organizes the complexity of multiple data sources into a manageable computational structure.
Data Source
Figure 1
Figure 2
AI summary
This disclosure relates generally to a system and method for product demand transfer estimation in retail. Demand transfer happens across one or more products within a category of substitutable products. Product level demand transfer indicates the amount of demand that is going to be transferred from the unavailable product to another product within the category. An inventory availability is considered herein rather than assortment to ensure that the product is present or not for a particular point in time. The system is configured to estimate product level demand transfer value based on the attributes values present in the one or more products considered. Moreover, the embodiments herein further provide product level demand transfer using the attribute value and demand transfer of values of those attributes that are different between two products.