Optimizing Retailer Supply Chain
A retailer’s shop is the final supply & inventory point, before meeting consumer demand. Its challenges are completely different than other types of the supply chain. Consider a grocery: It has many different products, with small amount of inventory, and pre-defined shipment periods. Mostly, there’s no chance to replenish the goods immediately. If the demand is not met, most probably the consumer is lost. In addition, the consumer can be lost forever, if it’s repeated. The demand prediction can be harder because of many parameters like seasonality, price changes, trends, promotions etc. On top of that, retailer’s cash flow is fragile and vital. He/she should not only maximize the revenue, but also balance cash flow.
Also, we have a very important problem, shelf capacity. Our grocery should display goods with many alternative, in appropriate amounts. We are not finished yet; Shelf Life, MPQ, Ordering Unit, Order Promotions and Quotas… So, we need to consider a lot of different things as constraint or limitation.
Under uncertain demand and above mentioned constraints, we need to optimize our small grocery’s supply chain. The Unified Supply Model supports this process very effectively.
In this article, we’ll start with basics and continue on others with some challenging issues, like shelf capacity, or revenue management.
First of all, we need to define the leadtime demand. We can apply the techniques given on Leadtime Demand Pattern. It can be good to consider the demand period as days. In order to calculate the leadtime, we need to check order & delivery mechanism. There are two common ways;
1) Hot-seller collects the order and delivers immediately.
2) Pre-seller collects the orders. Later, the transporter delivers.
For the first one, we just need to consider the hot-seller’s visiting period.
For second one, we need to consider the duration between order and second next delivery. Why? Because if something’s missing, we can order it on next pre-seller’s visit, and receive at next transporter’s visit.
Both of them have pros and cons. For example, the hot-seller case is better to decrease the leadtime, but some products may be missing in his van. Or, second one has longer leadtime, which increases the safety stock and order amount, but the shipment is under guarantee.
With applying the periodic demand (d) and leadtime (T), we reach to the leadtime demand pattern (D), and possibility function F(..).
Next one is definition of the optimum order quantity, (Q). The process is likely Optimizing Material Supply Chain, but with some differences. The order amount should be at minimum level to keep the inventory coverage profile close to the customer service level, CSL.
What about the safety stock? Before getting started at safety side, we need to calculate the Customer Service Level (CSL) of this product. For the inventory holding cost, we use the leadtime, which is calculated above. For the retailer, we don’t have back order cost, because near to 100%, consumer is lost. Therefore, it’s better to consider lost sales cost. Simply, it can be the difference between the price and cost of the product. We use it as (b) in our CSL formula.
To calculate the reorder point (R), we use the formula, given on Inventory Optimization. Here’s the simplistic version;
Avg[ F(R) + F(R+Q/2) + F(R+Q) ] = CSL
Avg[..] takes the average of the coverage profile of the inventory positions. Here, Q is the minimum order quantity, which is a product box (BX). We extract the reorder point (R) from this formula, with an iterative method. Whenever the physical inventory (I) is lower than R, we need to order Q, at least, to reach back to R+Q.
This is the simplest version of the Retailer Supply Chain application of Unified Supply Model. To reflect the reality, we need to add some constraints, like Shelf Capacity. I will continue it on the next article.