Fixing Root Causes

Fixing Root Causes Army Contracting Command was supporting U.S. Army Europe in 2006 as the latter worked to move soldiers and equipment to Iraq from European locations. Army Contract Command's mission was to supply goods and services to soldiers that supported mission accomplishment and provided for their well-being. The Command had some difficulty doing this due to the fact that its staff was spending excessive time generating Corrective Action Reports (resulting from errors in Command databases) rather than actually issuing contracts. A LSS team was tasked to determine how to shorten the time spent generating these reports. Instead, the team focused on fixing the root causes that drove the need for them in the first place. The number of reports was reduced by more than 50 percent and the Command saved $300,000 over four years.

Problem
Army Contracting Command (ACC) supports soldiers by procuring the goods and services they need to enable mission accomplishment and to provide for their general well-being. The Command provides contracting expertise needed to obtain these items while, at the same time, ensuring responsible stewardship of taxpayer money.

In 2006, ACC was supporting U.S. Army Europe in its mission to move soldiers and equipment from Europe to Iraq. Its support was hampered by the fact that staff resources were spending excessive downtime generating Corrective Action Reports (CARs.) These reports are required each time an issue with contract data accuracy or completeness of the ACC databases is detected, or an issue arises during issuance of a contract. After reviewing the amount of time spent generating CARs, ACC leadership decided to assign a Lean Six Sigma (LSS) Green Belt project team to resolve this problem. Personnel assigned to the team came from throughout the ACC organization.

Approach
The LSS Define, Measure, Analyze, Improve, Control (DMAIC) procedure was followed. During the Measure phase, the team quickly realized that the focus of this project should not be on shortening the time needed to generate the CARs as that would only improve a rework process. The team decided it had to address the root causes that drove CAR generation in the first place. The team started with a data sample from the CARs to identify the sources of errors. The data came from multiple ACC offices.

It was found that data accuracy (as measured by the number of opportunities for correct or complete data to be entered into the ACC database) was at a 3.6 Sigma level. (Sigma level indicates the reliability of a process and a higher number is always better. Sigma level 6.0 indicates near perfection.) While it was understood that 3.6 Sigma was acceptable in some manufacturing environments, the team knew that 4.5 Sigma was an industry standard for a transactional process such as issuing a contract. As a result, achieving 4.5 Sigma in the contracting process was considered an attainable project goal.

The team analyzed the data to find the sources of incorrect or missing information. It was determined that the issues were associated with the registration information in the Dun and Bradstreet and the Central Contractor Registration databases ACC used. These databases worked well in the U.S., but overseas, foreign resources were used for contract support. Analysis of the data found that foreign vendor information was the source of most incorrect data. The sources and types of defects associated with foreign vendors were found to be consistent across multiple ACC locations in Germany.

With the source of the defects understood, the team analyzed the process that foreign vendors used to enter information into the databases. The vendors had to register with ACC and in many cases, there were issues with vendors understanding the requirements for each entry field. The team reviewed the data to identify the fields with which the vendors had the most problems, and reviewed the instructions for information entry. The team used Voice of the Business requirements for each database field. The team also updated associated training materials and instructional information available on the ACC website. In addition to these improvements, the team added links to the website for vendors to use as "help" functions to further explain information requirements. Finally, the team updated internal training materials used for vendor meetings and seminars.

Results
After a one-month pilot of the new materials, the team measured a new Sigma quality level of 4.3 for new information accuracy. The team estimated the Sigma level would increase to 4.5 over time. The improvements significantly reduced the number of CARs generated by contract analysts by over 50 percent, and reduced required vendor rework of database information. This reduction in rework provided a cost savings of $300,000 over ACC's four-year budget cycle.