Sales Performance Deep Dive and Forecast: A ML Driven Analytics Solution
Download as PPTX, PDF0 likes403 views
This document summarizes a machine learning project to analyze sales performance and forecast sales for one of Unilever's declining brands. A team of data scientists used over 30 years of sales and market data to identify key drivers of sales and build forecasting models. Their best model achieved a 25% MAPE forecasting score by combining training and test data to account for the product's decline phase. Insights from the analysis will help Unilever's sales team improve business execution strategies.
Sales Performance Deep Dive and Forecast: A ML Driven Analytics Solution
- 1. Sales Performance Deep Dive and Forecast: A ML Driven Analytics Solution Team: Dijkastra Pranov Shobhan Mishra (VP – Risk Intelligenece at JPMorgan) Manav Tyagi (Solution Architect) Aniket Chhabra (Data Scientist at Paypal) Swapneel Gitccha (VP – Data Scientist at CitiBank) The codes can be found in the GitHub link below https://github.com/Pranov1984/Sales-performance-Deep-Dive-and-forecast----An-UNILEVER-Use-Case
- 2. AGENDA 1. EXECUTIVE UPDATE 2. BUSINESS OVERVIEW: PROBLEM UNDERSTANDING & ANALYSIS ARCHITECTURE 3. MODEL DEPLOYMENT ARCHITECTURE 4. MODELING OUTPUTS
- 3. PROBLEM STATEMENT • ONE OF UNILEVER’S BRANDS IS GOING THROUGH A STEEP DECLINE IN REVENUES AND IS REQUIRING MAJOR CHANGES IN BUSINESS EXECUTION PLANS. THE MANAGEMENT IS EXPECTING A THOROUGH ANALYSIS OF HISTORICAL PERFORMANCES CULMINATING IN IDENTIFICATION OF KEY FACTORS DRIVING SALES. DATA SUMMARY AND PRODUCT LIFE CYCLE OVERVIEW • THE DATA PROVIDED CONSTITUTED MORE THAN 30 YEARS OF INFORMATION OF SALES AND RELATED VARIABLES. • THE TRAINING DATA SUGGESTED THAT THE PRODUCT HAS GONE THROUGH A LIFE-CYCLE OF LAUNCH, GROWTH AND MATURITY. THERE WERE INDICATIONS OF A DECLINE PHASE IN THE LAST FEW PERIODS OF TRAINING DATA. • THE TEST DATA CORROBORATED THE INDICATIONS AS WE COULD NOTICE SHARP DECLINE (MORE THAN 25%) SINCE 2016. KEY INSIGHTS & DRIVER ANALYSIS • THE FACTORS HAVING A SIGNIFICANT POSITIVE IMPACT ON SALES VOLUMES WERE IDENTIFIED TO BE PROMOTION EXPENDITURE, VOLUMES PRODUCED OR IN STOCK, INFLATION, RAINFALL AND VISIBILITY THROUGH SOCIAL SEARCH IMPRESSIONS. • THE FACTORS HAVING A SIGNIFICANT NEGATIVE IMPACT ON SALES VOLUMES WERE IDENTIFIED TO BE BRAND EQUITY, COMPETITOR PRICES, FUEL PRICE AND DIGITAL IMPRESSIONS FORECASTING • MULTIPLE APPROACHES WERE ATTEMPTED INCLUDING ARIMA, HOLT WINTER’S DOUBLE EXPONENTIAL SMOOTHING, BAYESIAN APPROACH(BSTS) AND LSTM • THE BEST RESULTS WERE ACHIEVED WHEN TRAINING DATA WAS COMBINED WITH 2 YEARS OF TEST DATA TO CAPTURE THE DECLINE PHASES. MAPE OF 25% ACHIEVED WITH HOLT WINTER FOLLOWED BY ARIMA WITH A MAPE OF 33%. • FOR THE SECOND PROBLEM STATEMENT THAT REQUIRED TRAINING ON TEST DATA ONLY, BEST RESULTS WERE ACHIEVED THROUGH THE BSTS MODEL FOLLOWED BY LSTM. MAPES OF 5% AND 13% RESPECTIVELY WERE ACHIEVED. EXECUTIVE SUMMARY
- 4. • Build a process to estimate the sales forecast due to change in brand’s business execution strategy • Deep dive to the root cause of the factors contributing to brand sales • Leverage data, develop ML algorithm, provide sales intelligence using forecasting analysis to enable sales team to do their job easier BUSINESS OVERVIEW: PROBLEM UNDERSTANDING & REQUIREMENT GATHERING Background and Objective Data Elements Collect Marketing and Sales data and add dimensions Data Deep Dive Exploratory Data Analysis • Removing Inconsistencies • Univariate Analysis • Multivariate Analysis • Transformations Build Model Ready Data Sales Forecast Build an ML algorithm to estimate sales forecast Deployment Build framework to deploy model using technical stacks Analysis Architecture
- 5. WHY DO WE EXIST: AN ELEVATOR PITCH OUR MISSION IS TO HELP UNILEVER: • MAKE DISTINCTIVE AND SUBSTANTIAL IMPROVEMENTS IN BUSINESS USING DATA SCIENCE AND MACHINE LEARNING • TIE DATA TO BUSINESS DECISIONS THAT DRIVE LONG TERM STRATEGY AS WELL AS BUSINESS GROWTH
- 6. MODEL DEPLOYMENT ARCHITECTURE: AN AUTOMATED SOLUTION Input Data Processing Measurement Methodology & Modeling Pre-Processed data Financial KPIs Sales KPIs Marketing KPIs Customer KPI Method-1 Bayesian Time Series Model Forecast Model Results - Estimated Upside (Ensembled) Method-2 ML Algorithm Measure Pre- Post Growth Reporting Python/R Tableau Dashboard Reporting Upside Estimates Monthly / Quarterly Refresh Data Level • Promotions • Spends • Supply • Trade Promo • Price • Volume • Mkt Budget • Revenue • Categories • Engagement Implementation Tools
- 7. MODEL OUTPUT: ON TRAINING DATA (REQUIREMENT I)
- 8. MODEL OUTPUT: ON TEST DATA (REQUIREMENT II)
- 9. MODEL OUTPUT: IDENTIFICATION OF SIGNIFICANT VARIABLES