SAP SCM F&R Best Practices for Retail Industry

Abstract:

Forecasting & Replenishment (F&R) is the process of demand as well as replenishment planning generating order proposals and exceptions for the retail industry. SAP F&R is a solution meant for efficient replenishment of stores and distribution centers in retail industry. Its main goal is to simultaneously reduce inventory while optimizing actual service levels.

F&R is a multi-step process, focusing on forecasting for individual products and locations and replenishment planning based on forecast / reorder point and/or target stock. F&R focuses on products which are replenished regularly from external and/or internal vendors. Solution is capable to leverage quantitative techniques to forecast true market demand based on historical performance, past promotions, future events and demand generation activities. Solution requires the products to have a measurable and repeated sales history as well as up-to-date stock information to generate appropriate order proposals.

 

1.0 Business Process Definition

 

Forecasting & Replenishment (F&R) is the process of forecasting and replenishment planning generating order proposals and exceptions for the retail industry. F&R is a multi-step process and , focusing on forecasting for individual products and locations followed by replenishment planning based on forecast / reorder point and/or target stock.

F&R typically leverages quantitative techniques to forecast true market demand based on historical Performance, past promotions, future events and planned demand generation activities. Forecasting is followed by Replenishment Planning to generate order proposals and exceptions. Inventory Analysts review the order proposals and exceptions before releasing the order proposals for execution.

 

SAP F&R optimizes the internal logistics of retail companies by improving the replenishment processes and aims to achieve the following:

–Cut surplus stock in distribution centers and stores

–Reduce stock-outs in distribution centers and stores

–Lessen the large amount of manual work required by implementing highly automated replenishment planning in stores and distribution centers

–Increase transparency in the supply chain through effective analyses

–SAP F&R helps to minimize the total cost of ownership.

 

2.0 Forecasting & Replenishment Challenges Faced by Business

 

When clients focus on Forecasting & Replenishment for their DCs and Stores, they will face issues that could span supply, demand and even the nature of the products being forecasted.

 

Product Challenges:

• Large number of SKUs in retail business
• Capital investment products
• Inventory holding costs

Demand Challenges:

• Various patterns of demand for different SKUs
• Slow Runners with Intermittent Demand
• Requirement of tracking the demand signal daily

 

Replenishment Challenges:

• Various types of replenishment requirement for different SKUs
• Procurement challenges to satisfy vendor minimum order requirements vis-a-vis inventory optimization
• Manual procurement as well as smoothing of procurement cycles etc.

3.0 F&R Process Flow Overview

 

Following diagram depicts an overview of the business processes involved in the execution cycle of F&R.

 

 

To ensure a smooth execution of the above processes, business needs to adopt certain best practices with regard to the information management related to master data management, sale history and forecasting processes, replenishment processes as well as analysis of KPIs.

 

 

4.0 F&R Best Practices

 

4.1 Best Practices – Master Data Management

 

• Forecast horizon should be long enough to provide visibility for forecasting
• Master data changes performed in the ERP system, which are relevant for SAP F&R, need to be transferred to the SAP F&R system. This data is collected in a first step in temporary “buffer tables” in the SAP F&R inbound interfaces for later processing. Only after having received and processed all changed master data information in SAP F&R, the forecast and replenishment calculation process can be started. It is therefore important that the available time window for this process is kept and monitored. If the time window for sending changed master data information to SAP F&R is not sufficient the job for transferring the data can be scheduled more than once a day.
• Changes performed to transactional data in the ERP system, for example to purchase orders, promotions, contracts, stock and consumption data of F&R relevant article/site combinations need to be transferred to the SAP F&R system. This data is collected in a first step into the buffer tables in the SAP F&R inbound interfaces for later processing. It is therefore important that the available time window for this process is kept and monitored for accuracy.

 

4.2 Best Practices – Forecasting Process

 

• Forecasting should be done daily for all SKUs of all Stores / DCs
• Forecasting time buckets should be aligned with business requirements
• Forecasting can be done in multiple units of measure, including units, lbs and pallets

 

4.3 Best Practices – Historical Data Management

 

• At least three years of history should be used to form the basis for statistical and judgment forecasts, with the option to use more
• Demand history for forecasting models true Stores demand (requested dates and quantities), not actual DC Shipment history (e.g. late shipments, short shipments)
• History used for forecasting should be as close to the point of true customer demand as possible
• Lost sales should be systematically captured and applied to demand history
• History is adjusted for outliers that are detected and dispositioned based on standard rules
• Promotion history should be systematically captured and used to split history into base and promo demand
• Historical substitution, merging, and chaining should be enabled to support product transitions
• Daily Sales History stored in BI Cube causes Performance Issues for Conversion of Waves

 

4.4 Best Practices – Statistical Forecasting

 

• The starting point for all demand plans should be a statistical forecast available at the individual SKU or SKU/Store level
• Statistical forecasting models are established and periodically fine-tuned by a trained statistician
• A statistical forecasting model that generates the lowest error is systematically chosen
• Pick-best statistical forecast options can be limited by SKU to those that are appropriate to the demand pattern rather than all modeling options
• Manual Replenishment Type Changes need to be “Business-Rule-Based” and should not be decided by a “Situation”.
• Data for introduction of New Store & New SKU Algorithms should be captured from similar Sister Stores / SKUs.
• Listing period data plays an important role in forecasting of SKUs. Improper listing data would cause wrong seasonality resulting in over-stocking or under-stocking. Hence suitable governance  should be in place to maintain data related to listing.
• Documented comments capture the reason for significant variation from the statistical forecast recommendation.

4.5 Best Practices – Replenishment Process

 

• Vendor Lead time should be managed by Market instead of by SKU
• Improved KPI on In-Stock Position
• Data issues on Presentation Stock should be streamlined
• Data misinterpretation on Vendor Minimum vs. Rounding Profile vs. Min Order Qty should be managed
• Manual Replenishment Type Changes need to be “Business-Rule-Based” instead of “Situation-based”
• Restriction Profile should be setup for DCs on “Truck Load”
• Importance of Smoothing of Procurement Cycles across the entire business week should be emphasized
• Store Communication must be streamlined
• Manual Ordering – guidelines for Do’s & Don’ts should be established for better optimization of store inventory.

4.6 Best Practices – Performance Measurement

 

• Forecast accuracy and bias should be regularly reviewed and root causes of error should be discussed with forecast contributors
• Error calculations should be weighted to minimize the impact of large errors on low-volume and low-cost SKUs
• Lags for forecast accuracy measurement should be synchronized with and vary by product lead times
• Fine-tuning of system parameters for batch jobs is the key to a smooth functioning of daily planning process in F&R
• Setting of user parameters should be managed to avoid Performance Issues in F&R landscape.

5.0 Conclusion

 

Adherence to the best practices facilitates the smooth functioning of the planning process in F&R while optimizing the system performance and output of the planning algorithm. Data inconsistencies play a crucial role in F&R landscape. Unlike CIF in APO, where the transaction data are transferred from ECC to APO in real-time, buffer interface in F&R is asynchronous. Any synchronization between ECC and F&R is an on-going maintenance activity and needs to be ensured before every FRP run. While planning the data management strategy, planners must consider the performance issues due to data volume and schedule of the jobs to generate inconsistencies and their manual cleanup.

Best practices need to be established for respective organizations suiting to the specific needs of the business with regard to the master data management, forecasting process and application of demand influencing factors, replenishment process and optimization of order proposals followed by performance assessment of KPIs. Most importantly, bets practices should be treated as a set of guidelines and needs to be fine-tuned and optimized as the solution landscape as well as experience of the planners mature over a period of time.

 

 

6.0 Conclusion

 

Pravat Dash, CPIM is a Managing Consultant in the Business Consulting Services Group of IBM Global Services. He has over 16 years of SAP SCM/ERP implementation and industry experience in the area of Supply Chain Management and Logistics. He has worked as Lead Consultant and APO Team Lead for implementing SAP SCM solutions for clients in various industry verticals for the past 12 years. He has authored multiple papers in Supply Chain Planning space. You may reach him via email [email protected].