Monday, January 27, 2014

How Using Forecasting Model Drives Supply Chain Innovation.

How Using Forecasting Model Drives Supply Chain Innovation.  

Founded in 1962, Walmart now has 8,500 stores around the world. According to one study in 2006, 90% of the US population live within 15 minutes of a Walmart[i]. Moreover it is hard to talk about supply chain management without mentioning Walmart. In its relentless pursuit of low consumer prices, Walmart embraced technology to become an innovator in the way stores track inventory and restock their shelves, cutting costs and passing the savings along to customers.[ii] Walmart process of operation became synonymous with the concept of successful supply chain management.

 A key factor that Walmart can save money and sell product with low prices, it is because Walmart is only a retailer, not a manufacturer. Thus, as a retailer, Walmart uses the Logistic methods of cross-docking. It is similar with Zara. Those two companies which are Zara and Walmart aim to have items in inventory for a minimal amount of time, going from manufacturer to the shelves rapidly. By having items go almost directly from inbound to a warehouse to outbound to a store, it reduces unnecessary transportation costs and wasted time of items going through inventory.[iii] However as author Bill Sporito descripted Walmart has the trouble lurking on Walmart’s Empty Shelve.

 Thus, Walmart created forecasting model to fill up empty shelves to keep their customers. Within the eight months, Walmart made a $50 million swings in profitability by using this forecasting model.[iv] The legendary story of how Walmart profited from data sharing and how it improved logistics through better forecasting and inventory management is well understood; however, it has not been replicated to the same level by any other retailer to date. This can be attributed to the genius of Sam Walton, a Big Data analytics pioneer.[v]
There is an example how it works. For example, Supplier X might argue that Walmart should dedicate more shelf space to its products because of its high sales volume and high profit margin. Supplier Y would then crunch the numbers and argue that reducing shelf space of its brand might not seem so negative on its face, but that shoppers of its product also often buy additional products that carry high margins for Walmart. While Supplier X and Supplier Y jockeyed back and forth, providing greater insight with each analysis, Walmart gained a wealth of understanding about what was going on in the business. Each brought an alternative approach to forecasting, estimated their own and cross-price elasticity and shelf out-of-stock rates, calculated store and DC fill rates, analyzed assortment decisions, estimated inventory investment and cost analyses, derived return on investment for the shelf space, and illuminated category trends and its drivers.[vi]

How does forecasting model work?
Last year, my project group created forecasting model. This showed 86% of accuracy to forecast next 6 months. This process how to use forecasting model.
  1. Load Historical Data and Create Master Data
    Identify the key data elements that need to be considered and load them.
  2. Clean the Historical Data 
    There are almost always problems with the quality and completeness of the data loaded into the system. E.g. "demand" may not be true demand, because it is taken from "sales" data, and will not include "stock-outs".
  3. Generate a Statistical Forecast for Existing Products 
    Use demand planning software with built in statistical models to find a "best fit" that will give you a starting forecast.
  4. Prepare Forecasts for New Product Introductions (NPI)
    Use the demand planning software to identify products with similar sales trajectories which will be used as the starting forecasts for the NPIs.
  5. Override Statistical Forecasts with Judgmental Input 
    Use data from sales channels, knowledge about changes in market conditions, and expert insight to smooth the forecasts into the most realistic forecasts possible.
  6. Adjust the Baseline Forecasts for Promotions 
    In certain industries, like consumer goods, promotions can have a huge impact on sales volume and need to be factored into the baseline forecasts.
  7. Manage Vendor Managed Inventory (VMI) and Collaborative Planning, Forecasting and Replenishment (CPFR) Processes 
    Be sure to communicate data to both customers and internal managers responsible for these programs.
  8. Generate a "One Number" Forecast 
    Integrate forecasting into a Sales and Operations Planning (S&OP) that brings together executives from key areas of the company to ultimately agree on a single forecast number and execution plan that will drive both the demand and supply sides of the enterprise.[vii]

 Basically we used statistical regression equation and input some factors that influence sales. This factors is from big data. However, every factors cannot be used. It must test by regression model to see whether it has accuracy or not. It is true that my team’s forecasting model provided us a successful result. However I was still concerning about future uncertainty. There are many factors that we cannot expect. Because of this reason, my team could not increase it to 100%. I think Walmart will same problem with our group. Of course, it increased its profit up to a $50 million but to increase more profit, I think Walmart require to solve the factors that have uncertainty. Otherwise, Walmart can not grow more than this. 


No comments:

Post a Comment

Note: Only a member of this blog may post a comment.