AI Markdown Engine Using Clustering and Mixed Integer Programming
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Retailers face challenges in determining data-driven markdown strategies that result in either excessive discounts eroding profitability or excess inventory due to unscientific approaches, leading to out-of-stock warnings and increased costs.
Innovation Solution
A computer system with a markdown engine that optimizes markdown allocation across products using artificial intelligence and mixed integer programming, clustering products based on quantifiable relationships, and continuously learns to improve markdown plans through real-time sales data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If blanket promotions are applied across entire categories or regions, then inventory clearance is achieved, but profitability is eroded due to excessive discounts
Solution Approach 1:
The patent segments the blanket promotion approach by clustering products into distinct groups based on multiple attributes (category, margin, seasonality, demand patterns). Each cluster receives customized markdown strategies rather than uniform discounts, allowing inventory clearance while preserving profitability through targeted discounting only where necessary.
Solution Approach 2:
The system applies local quality by determining specific markdown percentages and timing for each product cluster based on its unique characteristics. High-margin products receive different treatment compared to low-margin products, and products with strong demand patterns are discounted differently from those with weak demand, optimizing both clearance and profitability locally for each product group.
2Reliability
If deep discounts are applied to clear inventory, then stock-out warnings are reduced, but excess costs remain due to too shallow discounts on other products
Solution Approach 1:
The system dynamically adjusts markdown strategies based on real-time and historical data about each product's demand patterns, seasonality, and margin characteristics. The optimization continuously adapts discount levels and timing to achieve the right balance between preventing stock-outs and minimizing clearance costs, rather than applying static deep discounts across all products.
Solution Approach 2:
The patent changes multiple parameters simultaneously (discount percentage, timing, duration) for different product clusters based on their specific attributes. By varying these parameters according to product characteristics such as margin, seasonality, and demand elasticity, the system achieves reliable stock availability while optimizing clearance costs for each cluster.
3Ease of operation
If unscientific approaches are used for markdown determination, then implementation simplicity is maintained, but measurement precision of optimal discounts is insufficient
Solution Approach 1:
The system performs self-service by automatically executing the entire markdown optimization process without manual intervention. The computer system clusters products, determines optimal markdown strategies, and generates implementation plans autonomously based on historical data and business rules, maintaining ease of operation while achieving high measurement precision through sophisticated algorithms.
Solution Approach 2:
The patent replaces manual, mechanical markdown determination with an automated computer-based system that uses data analysis and optimization algorithms. This substitution enables precise measurement and calculation of optimal discounts for each product cluster while simplifying the operational process, as the system handles all complex calculations and recommendations automatically.
4Loss of energy
If data-driven markdown optimization is implemented, then profitability is improved, but device complexity increases due to AI and mixed integer programming requirements
Solution Approach 1:
The patent introduces an intermediary layer (the computer-based optimization system) that bridges raw historical data and actionable markdown decisions. This intermediary performs clustering and mixed integer programming optimization to translate complex data patterns into clear, implementable recommendations, managing system complexity by encapsulating the computational complexity within a dedicated optimization module.
Solution Approach 2:
The system manages complexity by changing the approach from manual parameter adjustment to automated parameter optimization. The mixed integer programming model automatically determines optimal discount parameters (percentage, timing, duration) for each product cluster based on historical data, improving profitability while containing complexity within the automated optimization framework rather than requiring complex manual processes.
Data Source
AI summary
In an example implementation, a cloud computing platform receives input data including historical data representing a plurality of first items and a plurality of stores associated with the plurality of first items, and generates a plurality of first data structures. Further, the system uses an artificial intelligence (AI) system to adaptively cluster the first data structures according to item metrics into a plurality of clusters, and classify a plurality of second items into the plurality of clusters based on the centroids. Further, the system determines, using a predictive computer model, a markdown plan for at least one of the first items or the second items associated with that cluster, and optimizes the markdown plan using mix integer programming.