3.4 PER MILLION
Next in Line
Always look ahead to the next project for maximum quality gains
by Joseph D. Conklin
In quality, the sequel is what usually makes things better. Who knows how often savings have been left on the table because organizations failed to look ahead to the next project?
My friend Sam knows firsthand the value of lining up the next quality project before the first one is over. Just ask him about his job as a quality facilitator at Stream Shelter Research. One of his projects involved improving the visitor-request process.
"Why do you need a process for that? I just walk in or out the door, and that’s it," I said. It turned out not to be that simple.
Visiting the issue
Stream Shelter performs contract R&D for companies in its industry, a market subjected to government regulation. It takes ideas that pass proof of concept and helps determine the next step to full-scale production. The pace is fast, the business environment changes rapidly, and the array of customers is diverse.
Researchers, suppliers, potential customers and government auditors are always showing up to look things over, and that means contact with highly proprietary products and operations. Sam convinced me a process was required. He even showed me the request form (Figure 1) and flowchart (Online Figure 1).
A string of lost sales and canceled orders convinced management something needed fixing. In finding out why, one issue repeatedly mentioned was the cumbersome process for arranging visits. If it takes too long, customers lose interest, and audit deadlines for government regulators might be jeopardized. Errors in registration may cause authorized visitors to be turned away at the gate. If visitors are also customers, they rarely return. Media visitors, too, might not feel encouraged to offer Stream Shelter good press.
Management asked for the quality department’s help in improving the process, and Sam drew the assignment. He dug through the security department’s records for the last 12 months and estimated two important variables: the length of time to approve a visit and the percentage of visitor registrations performed in error.
When management saw the charts in Figures 2 and 3 for the first time, it was shocked and incited to take action.
Sam led the first improvement team, which flowcharted the visitor-request process, brainstormed possible causes of error, and developed the cause and effect diagram in Figure 4. The combination of the brainstorming and the cause and effect diagram led to the check sheet in Online Figure 2. The team wanted to know which of the possible causes were the actual ones. Thanks to the data revealed by the check sheets over several weeks, the team prepared the Pareto chart shown in Figure 5.
Turning things around?
When the team and Stream Shelter management saw the Pareto chart, the top three causes showed low-hanging fruit:
- "Pass not delivered" was traced to errors by the post office in reading handwritten addresses on the envelopes containing the passes. Security switched to computer-generated mailing labels to reduce this problem.
- Warped badge stock was traced to material deteriorating after being held too long in storage. Management purged the old stock and bought replacement material under a new policy of discarding material if it was not used by a certain date.
- Outdated badge readers were addressed by purchasing new ones and maintaining them more diligently.
The team continued to meet monthly to monitor progress. During the next 12 months, lead time and error percentages were gradually reduced to about two-thirds of their initial levels, as shown in Figures 6 and 7.
During these 12 months, Sam advocated to move past the low-hanging fruit and concentrate on material and equipment. The Pareto chart suggested the possibility of additional gains in the people and methods aspects. Sam warned that the less-tangible parts of the process might interact in such a way to limit the overall gains from the new material and equipment.
Attacking the less-tangible parts required a more holistic, integrated view of the visitor-request process. At the time, the process lacked an owner. Its parts were scattered across multiple functional groups at Stream Shelter. Under pressure from other urgent business matters, management concluded the tangible improvements were enough and eventually disbanded the team after it completed one year of monthly monitoring.
Not surprisingly, the performance plateaued in the next six months, as shown in Online Figures 3 and 4.
Please try again
The new status quo might have been sustained longer but for a combination of events that stimulated a second improvement effort. Even with the equipment improvements, enough disgruntled visitor stories made it back to the corporate office that management decided to compare Stream Shelter with its sister plants. Stream Shelter’s times remained the longest—twice as long on average compared to its sister facilities.
Meanwhile, Stream Shelter began preparations to achieve ISO 9001:2008 certification to meet government regulations and customer expectations. Because the goals of certification entailed removing self-contained functional silos as much as possible, the scattered responsibility for the visitor-request process across multiple departments felt at odds with the new effort.
When an untrained new hire failed to follow up on one important request, a major potential client was turned away at the gate because security personnel could not find approval for his visit. The resulting loss of a $50,000 contract immediately focused management’s attention.
Things grew really exciting when security rejected the request for the ISO 9001:2008 accreditation people the day before they were scheduled to arrive. The supervisor who signed the form no longer worked in Sam’s quality assurance group. A new request was expedited with noticeable encouragement and attention from Stream Shelter’s top management.
The shortcomings of the visitor-request process caught the attention of the accreditation people who independently seconded one of Sam’s standing suggestions: Start a second cross-functional team that included more operations employees, in addition to those from the supervisory and technical ranks.
With the benefit of increased management attention, the team received additional training in lean Six Sigma improvement techniques and benchmarking against sister facilities, and was able to assess several enhancements that, when implemented, encouraged a more cross-functional perspective. Ideally, the visitor-request process needed a single owner. Senior management identified a few potential candidates within the company. In all cases, a month or two was needed to transition responsibilities to free up the new process owner for the job.
Until an owner could be established, Stream Shelter management approved the second team’s recommendations, changed some of the team members and asked it to take charge of the initial implementation. The most important enhancements were:
- Assigning specific individuals as hand-off points among the requesting organizations, personnel and security to ensure a request reached the next stage of processing.
- Making visitor-request process training a mandatory completion item in new hires’ orientation. This ensured new employees did not receive their entrance badges until training was completed.
- Clarifying the approval procedure so security could accept the name of the supervisor who was in place on the date the approval form was signed.
- Removing the visitor approval list from the set of items requiring senior management sign-off for all changes. This had proved to be a barrier in disseminating the current list to all who needed it. Instead, senior management added administration of the approval list to the annual plant audit so the responsible employees could demonstrate they were handling it right.
- Rotating personnel between the two key departments—personnel and security—so both areas developed a cadre with more comprehensive knowledge of how to check the pertinent details of a pending visitor request.
Figures 8 and 9 show the progress in the 12 months after the second team started implementing its recommendations. By the end of one year, Stream Shelter’s visitor-request process performance aligned with its sister plants. At this point, Sam and all of Stream Shelter were proud and relieved at what had been accomplished. They were poised and prepared to make the visitor requisition process even better.
The column is based on a true story involving a real organization. Names and data have been changed to maintain the organization’s confidentiality.
Joseph D. Conklin is a mathematical statistician in Washington, D.C. He earned a master’s degree in statistics from Virginia Tech in Blacksburg and is a senior member of ASQ. Conklin is an ASQ-certified manager, engineer, auditor, reliability engineer and Six Sigma Black Belt.