Deep Diving in Retail Big Data to Excel Business Performance How IoT empowers BDA for Retail sector Kelvin Koo Business Development Manager kelvinkoo@clustertech.com +852 2655 6162 May 2015
Introduction of ClusterTech ClusterTech was founded in 2000. ClusterTech headquarters in the Hong Kong Science Park ClusterTech has 180 professional staffs with advanced degrees (16 Ph.D.s) in science and engineering, 4 of whom are/were professors. ClusterTech provides consultancy & implementation services on Business Intelligence, Data Mining, Social Media Analytics & Big Data Analytics for Retail, FSI & Gov t Utilities covering Asia Pacific Region.
Our Professional Team Consultancy Cornell University Ph.D. Stanford University Ph.D. Oxford University Ph.D. Michigan University MBA Kenneth CHOW Chief Business Consultant Francis NG Director of Business Ping Chung NG Business Consultant Jason WONG Business Consultant Data Scientist University of California, Berkeley Ph.D. Oxford University Ph.D. Chinese University of HK M.Phil. Chinese University of HK Ph.D. Russell YIU Chief Scientist The Chinese University of HK JD Chris CHOY Senior Computation Scientist Tony CHAN Senior Computation Scientist Kin Kwong LEUNG Senior Principal Scientist
Our Clients (Partial)
Our Expertise Big Data Infrastructure & Analytics Big Data Applications Pattern Recognition, e.g. Behavioral Modelling Constraint Optimization & Forecasting Retail & FSI Personalized Marketing Sales Forecast & Inventory optimization Gov t & Utility Big Data Infrastructure Data: Transaction data, POS, ERP, social network, sensors, unstructured data Simulation & Prediction / CLOUD Authorized Cloud Service Provider
Data Technologies for Retail Existing system Personalized Marketing Sales Forecast & Inventory Control
Personalized Marketing by Data Mining shopper attribute is the key Historical VIP records (CRM, POS, campaign record) -Product recommendation -Repeat purchase program All VIPs VIPs who bought product A but NOT product B VIPs who bought product B Member Classification Data Enrichment e.g. RFM, demographics, association, purchasing power, etc. Attribute tags VS Model development Testing Tuning High potential prospect! E.g. Decision tree, neural network, logistic regression, etc Campaign automation Marketing channels, e.g. edm, APP, WeChat, location-based promo, social adv, etc.
Sales Forecast & Inventory Control Key success driver for Omni Channel Regular demand prediction POS Probabilitistic Statistical Model, e.g. Poisson, GPEWMA Inventory Control System External factors Occasion demand prediction External Influence Seasonal Campaign Competitor action Clearance Change season Sales Impact Multi-factors time-series analysis, e.g. Regression Benefits: Speed up turnover Reduce OOS Max depletion Improve shopping exp Discover hidden factor influencing purchasing behavior
How IoT could help Modern Retailers to move their analytics to the next level? Big Data Analytics
Big Data Analytics Modern Retailers are experimenting IoT technologies bit by bit. They need a data collection & integration strategy.
Case study: Capture Mainland consumers wallet An end-to-end IoT + BDA + Personalized Marketing solution Big Data Analytics Photo source: http://jingdaily.com/
IoT applications, analytics & marketing executions are scattered In-store Touch points? WeChat Management platform Follower acquisition People traffic tracking WeChat marketing execution (e.g. e-coupon, referral program, retargeting, etc.) E-commerce site
Data integration for more accurate & actionable personalization Shopper Segmentation Shopping behavior attributes DataMart Browsing & purchase history Follower segmentation + Personalized message Click stream analysis Personal interest attributes Purchase behavior attributes Illustration In-store Reactivation Loyalty appreciation Purchase Analysis Offer recommendation
Case study: Service Excellence for Omni channel business Leverage IoT empowered analytics to enhance order fulfillment Big Data Analytics Photo source: dumblittleman.com
Omni channel business puts new challenges to order fulfillment excellence Mobile Wallet Modern Retailers are moving towards multiple delivery & fulfillment strategies: Click and collect Buy in-store, ship to home Ship-from-store Same day/on-demand delivery Market info transparency Photo source: thesocialworkplace.com; cues.org; healthblog.ncpa.org
More accurate & transparent inventory monitoring is required A dashboard to provide actionable insight: Coverage Holes Broken size Depletion Sales surge Slow moving And more.. Prioritized alert notice to selected buyers via Email SMS Auto recommend stock balancing actions Seamlessly integrated with stock balancing expert system
Agile inventory balancing is the key driver to Omni channel strategy Automatic inventory optimization system Unique algorithm to cater different product life cycle stages, e.g. fast moving, slow moving, change season, promotion, clear out, etc. En-routing Delivery planning Route planning Right Product in Right place at Right time Happy customer
Key take away: 1. Use IoT tech to collect shopper & inventory data 2. Choose a trusted data strategist to design the data architecture for a more complete shopper & inventory profile Big Data Analytics 3. Pick a quick win business case to experiment personalized marketing & O2O strategy 4. IoT tech is evolving, but it all boils down to data collection & integration strategy. It is not too late to start thinking about that!
Deep Diving in Retail Big Data to Excel Business Performance How IoT empowers BDA for Retail sector Q&A Kelvin Koo Business Development Manager kelvinkoo@clustertech.com +852 2655 6162 May 2015