Published on March 4, 2014
Planning for Complexity & the Predictability of Demand Sridhar Bashyam Director, Supply Chain Frito-Lay North America Ben YoKell Principal, Demand Planning Intelligence Consortium Chainalytics Supply Chain Planning & Forecasting Conference February 23-25, 2014 | Scottsdale, AZ 1
Agenda Company Overview Demand Planning at Frito-Lay North America Demand Planning Benchmarking 2
Agenda Company Overview Demand Planning at Frito-Lay North America Demand Planning Benchmarking 3
PepsiCo Overview 4
Who We Are PepsiCo is a global food and beverage powerhouse. Our broad range of more than 3,000 delicious products offers consumers convenient, nutritious and affordable options in nearly every country around the world. Global Beverages Global Snacks Global Nutrition Performance Brands Scale People more than 22 >200 ~300,000 billion-dollar brands countries & territories employees $65 billion revenue 5
Our History 6
From Seed to Shelf Frito-Lay owns the entire supply chain, so there are significant optimization opportunities. Move Information Before Inventory Raw Materials & Purchased Finished Goods Plant Warehouse Distribution Sales OPS Route Sales Retail Store R Product Flow Visibility to company wide information, enabling optimization and increasing the velocity through each function, while reducing the effort and improving productivity. 7
Our Integrated Supply Chain Statistically Calculated Safety Stock Multi-Echelon Inventory Optimization by JDA Software Production Planning & Scheduling Daily Production Planner by Frito-Lay North America Statistical Forecasting Models Execute & Deliver Base Forecast New Products + Promo Lift + Event Lift Cannibalized Products Discontinued Products Replenishment Planning Transaction Data SAP 8 Supply Chain Planner by JDA Software
Agenda Company Overview Demand Planning at Frito-Lay North America Demand Planning Benchmarking 9
The Whole Truth Every week we send a 100% accurate forecast through the system… …Sales just sells the wrong amount. 10
Demand Signal Threading in a DSD System 3+ Weeks Out 11 Week Of
FLNA Forecasting Facts Number of Regions 13 Avg. Number of SKUs ~800 Lowest Level Intersections ~8.5 Million Historical Sales & Promotion Data 3 Years Forecast Bucket Weekly Forecast Horizon 1 Year Forecast Frequency Once a Week Duration of Forecasting Process ~60 Hours Sales forecasted into the store R Plant 12 Distribution DC Route Sales Retail Store
Frito Lay DP SAS Solution Historical Sales & Causal Data Causal Data Current Functionality Promotion enrollments by account – – – SAS Promo Lift SAS Holiday Lift SAS Base Forecast Statistical Forecast Planner Adjust using DP UI Final Forecast Supply Chain Planning 13 Price point Duration Days in week Events: Holidays, quarter close, seasonal activities
Delivering the Final Forecast is a Collaborative Effort Product Supply Sales Input of accurate and timely promo enrollments Owns the signal Notification of non-price point driven activity Capacity and volume push Execution Region Finance Sales Ops Forecast feedback via weekly calls with replenishment planners Aligning DP forecast to the organization’s financial forecast Demand Planning Plants Forecast feedback from plant-based replenishment planners 14 Hold Key Account Managers accountable for accurate and timely promo enrollments
Demand Planning Benchmarking Benefits of Benchmarking Accuracy and bias benchmarking vs. relevant demand Use of demand segmentation and forecastability concepts – Region accuracy target setting – Assessing opportunities for accuracy improvements “How am I doing overall?” “How difficult is my portfolio to plan to for?” “How well should I be doing for my level of portfolio difficulty?” “How well could I be doing for my level of portfolio difficulty?” “Where should I focus my energy and attention?” 15
Agenda Company Overview Demand Planning at Frito-Lay North America Demand Planning Benchmarking 16
Demand Planning Intelligence Consortium Conventional Benchmarks Demand Planning Intelligence Consortium Questionnaire-based Data and analytics-based Participants self-report forecast accuracy however they measure it, which varies Reports are typically annual Level of aggregation is very high Limited comparative analysis Best practices ID’d without accounting for differences in portfolio forecastability 49% 33% PERSONAL CARE 13% HOME CARE 5% 17 FOOD & BEVERAGE PET CARE VS. Standardized transactional data collection and computational approach by unbiased 3rd party Reports are periodic and repeating Level of benchmarking is highly granular Provides understanding of drivers and causes Best practices ID’d accounting for differences in portfolio forecastability DPIC Statistics Industry: Consumer Packaged Goods Geography: North America and Europe Members: ~ 40 Businesses Item-Locations: ~ 500,000
Segmented Benchmarking & Analytics Nearly 1,250 Forecast Accuracy and Bias Benchmarks 18
How difficult is my environment to forecast? Segmentation provides an understanding of each portfolio, which drives demand planning performance potential as much (or more) as complexity. 19
How am I doing? Forecastability-Based Intelligence Forecast Accuracy Forecast Accuracy Conventional DPIC Forecastability Index™ Forecastability yields a very different conclusion! 20
How well could I be doing? Targets can be set with more realism by customizing for the specific demand portfolio, and should be differentiated: “One Size Does Not Fit All.” 21
Some Frito-Lay DPIC Findings Frito-Lay is a top performer, both overall and in segmented analytics results This is in contrast to assumptions around high complexity Top performance results stems from good process, great execution, and relatively high demand forecastability Forecastability + Process Design & Execution > Complexity Room still exists for improvement; even a top overall performer has segments of the business with opportunity 22
What’s Next for Frito-Lay Demand Planning Continue our membership in the consortium – Work on opportunity areas presented by survey results Add new causals to our SAS models Leverage point of sale data Continue to explore new technologies such as demand sensing Grow our modeling team into a Forecasting Center of Excellence (COE) 23
Thank you! Additional Information? Email DPIC@chainalytics.com Visit http://bit.ly/DPIC-info Connect http://bit.ly/ChainLI 24
... of the brightest minds in the industry. The Best of the Best S&OP Conference presented by APICS and the Institute of Business Forecasting and ...
Chaos theory; Command and Control Research Program; Complex systems; Complexity theory (disambiguation page) Constructal law; Cyclomatic complexity ...
Business Planning, Forecasting & S&OP Conference: Europe Amsterdam, Netherlands Register *This is a partial list. Other events ...
Planning of Construction Proj.pdf - Free ebook download as PDF File (.pdf), Text File (.txt) or read book online for free. Scribd is the world's largest ...
Supply Chain And Planning. Articles, experts, jobs, and more: get all the professional insights you need on LinkedIn. Sign up Get more personalized results ...
2016 Book-tax conformity and reporting behavior: A quasi-experiment by Evers, Maria Theresia & Meier, Ina & Nicolay, Katharina; 2016 Management of ...
Top JEL / L: Industrial Organization / / L1: Market Structure, Firm Strategy, and Market Performance / / / L11: Production, Pricing, and Market Structure ...
This article contributes to the related literature by empirically investigating the efficiency of nine energy and precious metal markets over the last decades,
THE NEED FOR INTERCONNECTION RESERVE The introduction and background above have recalled that one of the benefits of development of a large, interconnected