Graham Crackers
By Bob Boyles, Principal
Smarter Distribution

There is a theoretical difference between the way in which Gordon Graham wants you to manage your inventory and the methods that are evolving in the market place. The difference is one of manageability. The Graham model takes demand and calculates a "when to" order point and a "how much" to order quantity. These two factors come together and create an effective level of inventory turns, in essence each product is allowed to float to it's optimum inventory level. For most distributors this approach is a vast improvement over their existing manual methods of inventory management. However times are a changing. As with any endeavor there are a number of distributors that are seeking to improve on the industry standard. Some because they want to but most because they have to, we all know the reasons why, increased competition from industry consolidation, increased competition that from global competitors, the decreased margins that have resulted, reduced customer acceptance of shortcomings, changes in the distribution channel and changes in the links between wholesalers and manufacturers. The net result of this is increased pressure on the distributor to be more efficient. A number of distributors that are taking a second look at the Graham model and are finding it lacking in this more competitive time.

There are a number of areas that the Graham model has problems.

First, the Graham model does not allow the effective management of projected inventory.

Second, the Graham model assumes an unknown and intermittent demand level. Wholesaling is very different than a manufacturing environment where the demand is somewhat known and variance in demand is very small from period to period. Because of this the Graham model is structured to be reactionary, it looks at the past history and attempts to project the future sales that will take place. Because of this function it offers very little in the way of manageability. Most wholesalers are not comfortable with the "set the parameters and let it fly approach". The Graham model does not allow a distributor to target a desired level of inventory investment. Look at any MBA school text and the budgeting process starts with the sales forecast and the department budgets are derived from that number. By turning this logic on it's head the Graham model has robbed today's manager of one of the most essential tools.

Third, assumptions that the Graham model is based on are no longer true. The Graham model acknowledges that there are more complicated mathematical functions that can be used for inventory management however it states that they
should be avoided because the purchasing agent can't understand them. And what the average purchasing agent can't understand he can't manage. This is simply an outdated notion. When Gordon Graham wrote his book the personal
computer business was in it's infancy. Computation speed was still a precious commodity and networks were a thing of the future, much less a graphical user interface. System have evolved to where complicated mathematical formulas
can be calculated and then graphically displayed making even the most theoretical discussion mundane,

The Graham model for Safety Stock has a number of flaws. First, there is no mathematical way to tie the variance in the vendor's lead-time supply to the amount of safety stock. Also, there is nothing in the safety stock formula that takes into account the profitability or cost of an item.

The Graham model for Line Points has a number of flaws. There is no method to create a linkage between the customer's desired turns and the average stocking level. The idea of having to meet freight minimums that the Line Point formula is based on has gone out the window for vendors that are offering cooperative inventory management.

The Graham model for EOQ has a number of flaws. There is no linkage between line point and EOQ (they are very different mathematical functions). There is no inclusion in the EOQ formula of the profitability. There is no inclusion in the EOQ formula of vendor discounts.

Traditional inventory theory states that the goal of the purchasing agent is to balance inventory turns with stock outs. The Graham model does not allow for any type mathematical linkage between the two. The Graham model simply targets inventory turns.

In summary, it is time to take a look at some different mathematical models and use those to manage inventory.

About Bob Boyles and Smarter Distribution:  

Bob Boyles started his strategic consulting business in 2001 and has focused on the change that technology is forcing in the supply chain and how independent distributors can not only respond to that change but also maximize the return they are seeing on their investment. Bob has spent a significant amount of time as an Installation Consultant for several of the big name software companies in the distribution market. Working with hundreds of distributors across the country on installing, upgrading and utilizing their software.  Bob also worked as Corporate Systems Manager for one of the largest electrical wholesalers in the country as that company moved from a completely manual operation to an on-line real-time system.   

Bob is a graduate of Appalachian State University (BS - 1981)and University of North Carolina at Greensboro Graduate School of Business (MBA - 1985).  

© Copyright 2002, Robert S Boyles, Jr. All rights reserved. This article cannot be reprinted or reproduced in whole or in part, without the express written permission of Robert S Boyles, Jr.