The case is based on an insurance service company which traced back to 1994. The company was facing problems in controlling and improving service quality and operation performance. This essay will stress on introducing the SPC (statistical process control) to DAV and providing suggestions on improving both the SPC and the company’s operation performance. The five questions required are covered and allocated into different parts of the essay.
DAV: Situations and Problems
1. Background Information
1) DAV was one of the world’s largest insurance companies whose main business is in Germany and in retail insurance. DAV knew that it was significant to maintain that position as smaller insurance companies armed themselves to attack their primary market and were seeking the way to improve the management process.
2) DAKG was the back office part of the operation, focusing on the retail side of the business and it processed applications from new customers for policies, changed policy information, carried out various legal registrations, and kept track of a customer’s personal information.
2. Problems and current methods
Problems: As DAV’s Business volumes were exposed to seasonal fluctuations and longer circle change in the economy, two critical problems emerged as a drawback to improve the operation performance:
1) A difficult capacity management problem emerged as in the heavy days the transaction volume will be twice of lighter days. The work quality decrease significantly during the heavy days; meanwhile, in order to monitor and control the quality, tremendous human resources (and possibly highly unpracticed and inexperienced) were raised.
2) In measuring and controlling the process’s quality, DAV was using accuracy rates of different processes to get the information. However, the drawback of this method is that it only gets “numbers”, but the issue of “what to do with the number when you get it” cannot be solved. As the previous simple method of fraction defective average is not enough for operation management – Having only fraction defective on hand is meaningless, solving this problem will be the key in improving the operation performance.
3. What DAV needs
In conclusion, what the DAV need is a project as Process Measurement and Improvements (PMV) that can effectively monitor and control the process’s quality. What is important is introducing an instrument/ tool that can appropriately measure and reflect the process accuracy rate.
SPC: Introduction and Sample
To introduce the SPC to DAV (in a 30 minutes’ presentation), we will show the SPC’s definition, its advantage and fitness into DAV comparing to the previous method, together with a sample to illustrate how it works and controls.
1. SPC: Definition, advantage and its primary challenges
Statistical Process Control (SPC) involves using statistical techniques to measure and analyze the variation in processes. It is most often used to monitor the consistency of manufacturing processes. The goal of SPC is to get and keep processes under control.
SPC has a considerable advantage of early detection of problems. Under SPC, variations in the process that may have negative impact on the quality of product or service can be detected and corrected. SPC is not only about the numbers, but also about the checks on the process that whether the variations are out of control. SPC includes control charts like X-bar Chart, R-Chart and P-Chart that represent certain descriptive statistics. P-chart of SPC is a very important control chart for quality attribute data. It can measure the proportion of defectives or nonconforming units.
In this case, SPC is a considerably better method for DAV than its previous simple average method to improve its customer service operation, because it provides consistency in quality control and benchmarks for managers to verify the process. The average of fraction defective is merely a number which cannot tell us what to do with it even when we get it. Thus SPC is more accurate and systematic to analyze the process of customer service operation.
2. Primary challenges
There are basically two challenges that implementing the SPC to a service company:
1) Difficulties of deciding and selecting what to measure and
2) Too much uncertainty because of the human behavior
3. Sample: how SPC works
Here is an example of employing SPC (P-Chart) for DAV’s PMV project.
The first 12 weeks of the data displayed in Appendix I represent the diagnostic period for the Policy Extension Group.
From the calculation (See Appendix I), we get the 3-sigma control limits for the process, i.e. UCL=0.091, LCL=0.014. These control limits indicate that if the error proportion is within the range of [0.014, 0.091], the process is under control; if not, the process is out of control.
Here is the P-Chart for the data from Week 13 to Week 30.
It is then clear that the process is out of control in Week 23 and Week 24.
In addition to process control, SPC can also help lower rank employees to better understand the different standards of the services and draw their attention to self-control of fraction defective in their own performance by easily using benchmarks from P-Chart.
SPC can facilitate the process measurement and improvement, but some challenges still exist in applying SPC to a service industry compared with manufacturing. One of the primary challenges is that it is difficult to determine what variation is used to measure. Different groups in DAV may have varied kinds of standards to appraise their performance, and each group may include several variations to judge its own performance as well. Another primary challenge is that the benchmark for good or bad judgment is vague to decide. There are no definite standards like those in manufacturing industries. Despite the challenges, SPC is still a very effective way for DAV to control and improve the operation process.
Improvements: how SPC can fit into DAV
As the SPC is initially developed for manufacturing process control, before introduced to DAV, some important issues and improvements should be marked out. From the feedback of the two-month’ experiment, improvements can be summarized as following:
1. It seems to be contradictory that the better-performance teams have to do more sampling. However, as long as the managers and employees understand that the objective of the sampling is to draw a P-chart to detect errors, there should be less confuse on this issue. The managers should be reminded that ultimate goal is not to compare the accuracy, but to find errors and mistakes to further improvements.
2. It is much more difficult to define a mistake in the services company than the manufacturing process. The services’ quality is difficult to be objectively indexed to be measurable. However, if the managers can develop and allocate different weight to different items in the services process, the definition of a mistake will be much clearer. For example, they can say an error in recording the customers’ address will be valued at 1%, while the error in recording the customers’ contact will be valued at 2%. The managers can then set up a materiality level, for instance 5%, to be the minimum total errors that will be viewed as a mistake and important.
3. Measuring the lawyers will be another difficult issue, since the lawyers’ work accuracy is highly un-measurable. However, if we evaluate the service upon the effective and efficiency from the management perspective, we can still develop an appropriate way to use the SPC: on one side, we can ask the external customers to give a feedback on a rating scale to evaluate the service’s quality; on the other side, we can ask the internal supervisor to give a feedback on a rating scale to evaluate the lawyers’ effectiveness and efficiency. Combining the two different sources of rating will provide a good prospective in defining a “good” legal work.
4. It will significantly reduce the manual work in recording and tracking the accuracy if some work can be automatic charted. But the accuracy can be charted only if the variables are measurable in quantity. And what should be stressed here is that we should use the difference between the actual and estimated figure to evaluate accuracy. For example, as the problem solution team, the supervisors can preset an estimated time for process, based on their experience on similar works. If the actual time to finish the job exceed a certain percentage, we can then view the work as a “defect” or “mistake”.
5. For managers worrying about the misunderstanding in the P-chart, there will be some solutions to this issue. First, the directors will be trained to have a basic knowledge of the P-chart and its functions. The directors should understand that different services group/departments will have different measures to generate a P-chart. Secondly, regular report will be provided to the directors which stressed on the change of the accuracy, rather than single one period’s figure. Meanwhile, the major function of the p-chart is to control the accuracy rate in a certain range and identifying any abnormal figures and potential mistakes.
Further concerns: How to improve DAV’s operation performance
The purpose of SPC is more than to know what exactly the accuracy is. The ultimate objective is to improve the accuracy of the process. So we will step into the second phase to improve the performance using new measurement as the basis.
By analyzing the tested data, in this case which is, the proportion of error, we need to identify the factors that cause unusual data. Take the sample data above for example, why the process is out of control in week 23 and week 24(The case mentions the problem in heavy days, suggesting this may be the reason for fluctuation in short term.) Once we identify the factor we could try to eliminate or reduce the factor effect, which involves the improvement of operation process.
These are some possible process improvement:
1. The double-key entry method used may indeed improve the accuracy of transcription, but it involves the trade-off between quality and cost. Actually, it makes sense that the first-pass quality deteriorated as a result of double-key. So if first-pass accuracy can be improved to an accepted level, the double-key method can be abandoned in some conditions.
2. 12% of policy application forms are not in good order, leading to extra 20 minutes to process it. Although the case mentions that designing forms with clear instructions did not work well, we suggest that an appropriate printed sample can be shown to customers. We will identify the information easy to be missed or misunderstood by customers and we emphasize such contents in the sample.
3. We suggest that an application system based on Internet can be applied. With the system, customers can print information on the website of DAKG and pay online. If any information missed or printed in incorrect way customers can not submit the application until it is corrected. In fact this online process is applied widely today. This new process can reduce both the extra cost mentioned above. First, the system will inspect the accuracy of information and, second, the associates only need to review the application form which will then be transferred to database. It is much more effective and efficient, if the technology limit not taken into considered in this case.
After the improvement of process is applied, we need to use SPC to track the effect of new process. Also we should use data to diagnose whether the new process is under control, thus then identify the new factors functioning.