Amazon.com “Buy Box”: Getting Incremental Revenue through qualifying for the “Buy Box” by using an “Integrated Early Warning Solution” and Deep Learning Approach.

Amazon.com’s “Buy Box”:Its the box on a product detail page where customers can begin the purchasing process by adding items to their shopping carts. When a customer buys any product from Amazon.com’s website, he/she instinctively clicks on “Add to Cart / Buy Now” button. This tiny box which has a particular vendor is being called as “Buy Box qualified Vendor”. Several vendors compete with each other to qualify.The Buy box is very important to Amazon.com and the vendor because over 80% of the Amazon.com’s revenue is being generated through it and without getting qualified as a Buy box, a vendor loses its share to its competitor.Fast track is the commitment from Amazon.com on when exactly the product can be delivered to the Customer. Combining the Fast track and Buy box, Amazon.com has created a metric called “Fast track Buy Box”, which is the most important metric to measure the success of a vendor.

On several instances, Hewlett Packard was not qualified as a “Buy Box Vendor” due to lack of Inventory and so lost revenue to its competitor. There was a need to 1). Develop an Integrated closed loop solution which provides a comprehensive view of Inventory position, Order pattern, fulfillment status etc. using various data from Amazon Retail Analytics (ARA), HP SAP Systems etc, 2). To predict the order qty from Amazon.com on HPI(Hewlett Packard). In this paper we propose an “Integrated Early Warning System”, which helps to identify the pattern which leads to Out of Stock situations and Deep Learning approach (Long Short term Memory, special case of Recurrent Neural Network ) to predict the Orders placed on HPI to enable HPI to fulfill the orders and take pre-emptive actions to avoid stock-out situations.The total Incremental orders through this System generated is $ 8 Mil for FY’18.

 
4 favorite thumb_down thumb_up 0 comments visibility_off  Remove from Watchlist visibility  Add to Watchlist
 

Outline/Structure of the Case Study

  1. Problem Statement
  2. Reasons for not qualifying for Buy Box
  3. Approach to the to solve the problem using "Early warning system" & "Deep Learning"
  4. Model Framework
  5. Results: Feedback from Amazon.com
  6. Future steps

Learning Outcome

1. Identifying the patterns which lead to Out-Of stock incidents.

2.Application of Recurrent Neural Network to predict Amazon.com's order o HP.

3.Deciphering Amazon's "Black box" algorithm to select the Buy box vendor.

Target Audience

Data Scientist, Forecasters, Online Retailers

Prerequisites for Attendees

The attendees should have basic understanding about

  1. Machine learning/ Deep Learning
  2. The concept of "Buy Box" w.r.t Amazon.com sales
  3. Online retail strategy
  4. Demand Planning & Forecasting
schedule Submitted 1 week ago

Public Feedback

comment Suggest improvements to the Speaker

  • Liked Aswin Nair
    keyboard_arrow_down

    Aswin Nair / Raj Mohan - Attribute Based Component Forecasting for High Technology Industry using Machine Learning

    45 Mins
    Case Study
    Intermediate

    The Personal System Business Unit, the flagship unit of HP Inc, is powered by some of the most innovative technologies in the industry and has consistently delivered exceptional results. However, the business has been plagued with recurring shortages and over stocking of slow-moving SKUs & Components owing to poor forecast accuracy. The current forecasting framework uses conventional forecasting methods and basic time series models to arrive at the baseline forecast. This approach works well for certain segments and regions with high predictability and noticeable seasonality but fails for areas with erratic demand and weak seasonal behavior. This created a need for HP to develop a robust forecasting approach/framework to improve the accuracy. In this paper, we propose “Attribute based forecasting framework”, a multi model solution, which uses techniques like Text Analytics, Decision Tree, Random Forest, Support Vector Machine and Artificial Neural Networks in building the prediction models.