Presentation by Dhruv Boruah at Magento Live UK http://magentolive.com/uk/schedule/
Text of Predicting Retail KPIs using Magento & Machine Learning
machine learning Magento &
About me
Dhruv Boruah Co Founder, Reckless.io
Agenda
Agenda Your KPIs Session Goal Interesting Magento Data Magento Case Tools/How? Demo Q&A
Your KPIs Are you asking the right questions? How much do you know about your customers? AOV/LTV/GPM Sales/Returns/Shippin g Churn Rate Abandonment Rate External Factors Weather/ Forex/ Holiday/Newsletters Census (birth/death/age)
Machine Learning A branch of artificial intelligence, concerns the construction and study of systems that can learn from data. Source: http://en.wikipedia.org/wiki/Machine_learning
Session Goal
How to predict customer behavior using machine learning?
Interesting Magento Data
1. log_visitor_online (visitor_id, type, ip, customer_id, last_url ) 2. log_visitor (visitor_id, session_id, first & last visit, last_url, store)
Case Retailer A trades in: Clothing/ Toys/Shoes/Home etc. Goal: Predict checkout intent of online customers Deliver personalize promotion to prevent cart abandonment
Tools
How?
How? Listens for events Does delta sync Processes responses Magento Extension Scalable Infrastructure Slave Database NoSQL Google/Amazon
How? YES/ NO Slave Database Flatten Magento Tables NoSql database Data Warehouse Transaction History Merchandising Operational Marketing Modeling Preparation Training Predict Predict
Training Model Successful Checkouts Created date/time Remote IP Base subtotal Items Customer is guest Category Location Training Model Unsuccessful Checkouts
