Summary by Carolina Saavedra
Master of Accountancy Program
University of South Florida, Summer 2002

ABC Main Page | ABM Main Page | Cost Management Main Page

This article describes what happened to one of Hewlett-Packard’s divisions, the Colorado Springs Division, when they tried to implement ABC to provide better product costing and inventory valuation and had less than effective results.

Hewlett Packard, a decentralized organizational structure, had previously had other divisions successfully implement ABC. The Colorado Springs division within the Testing Measurement Group was a larger division that had a high-mix, low volume product line when compared to the other divisions. Their product line tested and measured equipment marketed primarily to design engineers in the communication and computer industries. The Testing Measurement Group management decided to implement ABC in that division in order to get better cost information on which to base product pricing decisions. The first step they took was to fully understand their processes and to identify the cost drivers. Then, its information technology group and its accounting group set up and modified ways to track the drivers through the computer systems (applications). One example of this is how they modified the materials’ computer application to identify preferred parts vs. non preferred parts as procurement drivers. To determine the preference priority, five criteria were used: technology, quality, reliability, deliverability, and cost. Based on these assessments they assigned parts to either the preferred, neutral, or non-preferred category. By doing this, R&D was able to have input regarding the parts assessment. It was important to distinguish between preferences because each part category had a different overhead cost rate to help influence the part selection process. This “spider web” information matrix required a lot of employees to plan and implement. The division used finance personnel, R&D, marketing, procurement, and manufacturing areas to help identify the cost drivers.

There were four reasons why ABC did not work for the Colorado Springs Division: 1) there were too many drivers 2) no proper administration 3) no follow through 4) and too much emphasis on consensus.

                 1) The participants at Colorado Springs selected drivers for almost every activity in the process instead of picking drivers for just the top two or three critical activities of a particular process.  They ended up identifying drivers and setting driver targets for a huge number of operations. They had over 20 cost drivers for just one of the process areas in manufacturing which resulted in an extremely large information matrix. Management should have established priority goals, focused on the critical processes, and then decreased the costs associated with those processes. By identifying and understanding the critical activities that drive the costs you are better able to cut down costs like labor, materials, and overhead.

                2) The Division began targeting drivers - forecasting overhead consumption and cost driver utilization. They also started tracking, documenting, and explaining all the cost driver differences on a monthly basis to include efficiency variances, rate variances and volume variances for the drivers. It simultaneously drove an overhead structure in the administrative area that it could not have supported. You cannot add value added benefits to a cost structure by explaining driver variances over a short period of time when the cost structure is relatively fixed over that period of time.

                3) The follow through needed to take advantage of the cost driver information was unclear, inconsistent, and not persistent. The division was too focused on the monthly variances that it was never able to follow through in delivering a cost reduction. There was not enough time for the Division to manage the overall cost structure much less provide for discussions between the cost driver creators (OH demander) and the users (OH consumer) which should have been essential. The choices made by the cost driver creator (such as R&D) had no effect on the users (such as procurement) cost structure when it should have. To fix this follow-through problem, they should have clearly defined goals, set workable metrics, strived for overall reduction of driver volumes in non-preferred areas, and reduced the overall cost structure.

                4)  Given Hewlett-Packard’s decentralized organizational structure, the division should not have placed so much emphasis on trying to organize cost pools on a consistent basis across all of the divisions. Although they tried to achieve this consensus to be able to use cost drivers as a benchmarking tool and to provide comparable measurements across the groups divisions, it was not a good idea since each division consumes overhead very differently either due to their strategies, management styles, or organizational differences. Divisions should not benchmark at the cost driver level but instead at the manufacturing overhead or gross margin level for similar business units.

In conclusion, ABC allows a company to gain an understanding of its cost structure from a process perspective and to find the real value of its product often in the form of inventory. For Colorado Springs Division on the other hand, it helped them understand its manufacturing processes, overhead consumption, and manufacturing efficiency, but it did not help them with their inventory valuation method. ABC was a good continuous process improvement measure because it identified the processes that needed improvement and was an invaluable tool for make/buy decisions since it provided an understanding of the cost structures and the activities that drove those costs. It also required that the accounting department, manufacturing, and R&D work together in designing improved manufacturing processes. Inventory valuation failed because it was difficult with the high heterogeneous product mix to try to understand all of the processes and cost drivers and deviations. This division also had high fixed costs (85%-86%) so ABC couldn’t have a short term impact with regard to product costing. For them it would be better to focus longer term in using ABC and cost drivers to find their core competencies. Another problem was that in order to monitor the driver activities and driver volumes, and actual versus targets, and the spending  the  accounting system was bogged down. There was more focus on manipulating the data rather than using the information to help the overall system.

ABC can be a great communication tool as long as there is a strong partnership between the different departments such as R&D and manufacturing. It can drive empowerment if everyone understands their role as service providers as well as understand who and what drives the cost structure of their service. In order to improve gross margins, you need to understand the cost structure of the organization and some key cost drivers and then communicate that information to the customers or partners of the services or products. ABC then steps in to help the customers or partners change their processes to improve their cost structure. In order to have success when implementing ABC you need to have clear goals, simple tracking methods, close internal and external customer collaboration, and consistent and persistent follow-through.