Deutsche Allgemeinversicherung (DAV) is one of the world's largest insurance companies. In 1996, 51% of DAV's business was in Germany (in which 60% was in retail insurance). Managers of other firms say DAV's strengths lie in both 'sound, traditional insurance management' (most likely the core insurance product offering) and 'outstanding customer services' (most likely the human contract element in the insurance product). The insurance company has cutting-edge technology. DAV seeks to lower its defects' internal costs and external costs by increasing its appraisal and preventation costs.
Concern#1: Insurance service products are increasingly becoming homogeneous (and easy to replicate), and not only does DAV compete against the other giants of the industry such as Allianz, Credit Lyonnais and Aetna, but in addition the company has many smaller insurance companies as competitors. Differentiating DAV's customer service can be used as an advantage over competitors.
In the case quality for customers is what customers value, i.e.: polite employees on the phone or through letters who can do their job rapidly without problems. It is important to think from the customer's perspective. It is very significant as this small attribute can give DAV a competitive advantage.
Concern#2: Specifically what customers value as quality and what they expect can be further researched through questionnaires or feedback. Quality is difficult to be measured in service industries due to the human element which has inconsistencies.
Concern#3: Delivering quality to customers posed a colossal mission for DAV to achieve. The case says that DAV had difficulty in providing customers with consistent, excellent quality as operations were run in many divisions in different locations. The size and diversity of operations made providing constant quality a very difficult task to achieve. A very significant part of this concerns processing data without mistakes and to be able to quickly access this information, development was necessary in this area.
The customer service group at DAV is known as Kundendienstgruppe (DAKG). DAKG focuses on the retail side (applications, policies, legal, personal information) of the business. According to Kluck it is a 'high-volume production envrionment'.
It employed 2,000 employees at 3 sites: Munich, Koln and Hamburg who operate in one of seven divisions of the company (page 3).
Concern#4: DAV uses temporary staff in addition to its permanent staff to reduce costs due to seasonal changes. This turnover might add to inconsistencies in customer service quality. Are temporary employees provided with sufficient data-entry preparation?
Concern#5: 'same-day' processing in general sounds appealing. However despite cutting edge technology and the extra temporary workers; if employees work longer hours it is detrimental for the long run as employees are likely to make mistakes when they work longer hours. There is room for improvement here.
Cannot replace the whole system by automation as some customers have unreadable handwriting or forget to add information such as their address.
New policy setup application form process
- Once a customer has filled in an application form they give or send it to one of eighty DAV branch offices.
- The branch offices send the form to the VEG division in Hamburg.
- Associates at VEG scan and process application, enter information into their database.
- If the information is not complete customers are called by another associate. 12% of application forms are not in good order and take on average an additional 20min to process. Customers are automatically sent a letter if their information is complete.
- If a customer sees a fault in their details, they can contact the company through a toll free number or a letter. The customer problem resolution department (in Munich or/and Koln) deals with this. Concern#6: Problems here were often due to incorrect entries into the database by associates. This causes huge frustration among customers. Although it may take only 5min for customers to contact the company, it would take an associate over 1 hour to correct the entry in the database.
The PMV project
Kluck was the chief of operations at DAV is behind process measurement and improvement (also known as Porzessmessung und Verbesserung, or PMV). The PMV project was 'revolutionary' as it involved applying manufacturing -style improvement methods to insurance services. When the employees make the checklists in the PMV about what is relevant to measure and what isn't which makes them think about priorities in processes. The PMV project had 2 main phases:
1) Measure the quality of a number of process steps (such as transcribing information from application form to computer). According to the case, this phase is completed, and the company is tracking performance over time. Concern#7: was this done well enough?
2) Use the measurements of phase 1 to improve performance of a range of processes. Phase 2 has just begun and Kluck is facing big difficulties with this. Concern#8: what should she do?
Statistical Process Control (SPC)
The application of statistical techniques is to decide whether a process is delivering what the customer needs. SPC is an integral part of Total Quality Management and Six Sigma. SPC is usually used by companies in the manufacturing industry. Insurance companies are in the service industry, where problems are harder to evaluate due to the human element and defining what to measure is more difficult. SPC is good at measuring processes such as the process of inputting data from the application form into computers and uses a p-chart (exhibit 2 in the case to use this).
Schoss pointed out the importance that SPC measures process rather than the people in it. He said that associates need a tool, in this case it would be 'quality management', and thirdly he emphasised that employees need to be protected by the fury of senior managers. SPC uses control charts (Upper and Lower limit, and a mean). DAV had technical issues as they did not know what a right size sample to use. Each application took on average 5minutes to check. It is very necessary for them to define the total process flow, critical process parameters (criteria), the sample size (300 to allow for 3 wrong answers?), over how much time will this take place (8weeks?), create a comprehensive automated charting system, allow room for correction, spot ways of development.
Kluck hired a consultant named Kober to help with training and HRM.
Concern#9: Soon after, 15 departments were involved in the experiment, are they all necessary?
Concern#10: Samples were set across the board at 200/week. Some groups which have less accurate processes had excessive sampling.
Proposed solutions for measurement challenges
1. Better teams do more sampling
From the start, 15 departments and processes are too many to measure. Kluck should focus on the new policy set up division before measuring others. If it is successful, only then should they use them on other departments. In the case the 99% accuracy figure is correct and we are to measure it if there are to be at least 3errors, the sample size must be at least 300. A measurement time period must also be included.
2. When is a mistake not a mistake? When it's not important
Guidelines such as what is an error and what is not one should be defined before measuring. An example of this is where inaccurate phone numbers are not important (as they are rarely used) but incorrectly entered addresses are very significant.
3. Measuring lawyers
The legal department had difficulty deciding what to measure. Perhaps customer evaluation forms can be used instead of the SPC experiment. Greater emphasis should be placed on important issues, rather than trivial ones, perhaps the real concern isn't a yes or no straightforward answer but needs some explaining.
4. Automatic charting
Divisions that have clear cut processes in which attributes can be clearly defined as right or wrong parameters can use automatic charts.
5. On the prowl
Although it is natural for managers to react in the negative ways they do, after seeing real measurements they need to be explained what they should do to manage their employees better, and that bad behaviour on their behalf is only unfavourable to the company.
1. Interview customers
This would be done to ask them what they value, and why they would pick one insurance company's products over another. This can be done either through focus groups or questionnaires.
2. Possibly move all employees from 3 sites to 1 main site.
If there is one main site instead of three there is likely to be better communication within DAV, and the employees are likely to possibly benefit from an increase in economies of scale and scope, and inconsistencies are likely to be reduced. This will be beneficial in the long run.
3. Interview employees
Although only senior managers have been addressed, there is no mention about how the employees themselves feel apart from the overtimes and the fact they are very busy. There was an excited atmosphere but have the employees been properly explained the role and outcome goal of the research? Have employees questions been addressed? Employees would be more willing to provide accurate data if they were told, perhaps this would also give them a slight incentive to be more careful at work.
4. Six Sigma Black Belt
It would be in the best interest to DAV to hire a Six Sigma Black Belt who specialises in improving quality of the project.
5. Working time
Employees should not work more hours than they can handle and this can be solved through different shifts among employees either from permanent or temporary.
Managers and their departments should develop realistic checklists to make sure everyone is working as best as they can, and to train those who are behind on schedule.
Improving performance phase (part 2 of PMV)
Analysing and evaluating data
It is important to correctly see positive or negative results in measurement. It is necessary to recognize real causes, distinguish ways of improvement and spot solutions. Once the causes have been identified managers should act accordingly. Where a point on a p-chart (for example in exhibit 2 in the case) is can be for many reasons so it is important that the assumed cause is tested out and seen, in order to make any valuable steps forward to improve the process.
It is important to monitor the performance of departments but also employees based on cost and time efficiency. Especially the ones who are not performing well would need training (temporary workers might not need a lot of training as they are not permanent-training them extensively might be a wasted investment). It is important for departments to have improved communication between one another, this would allow better collaboration, encourage competitiveness, and allow each employee to understand their position better in the company. Employees will also learn how to carry out processes better through this. Measuring performance over time must be an on-going process and someone needs to question what numbers mean comparing performance figures depending on dates (seasonality and any holiday must be taken into account).
By using checklists in departments, one can find out the source of the cause of progress or a problem. This can either be used to an advantage that should be made an example of or should be ratified. Processes in the company are linked and one problem in one process can have an effect on another process (so the error might not be evident at first glance), therefore employees must keep in mind the organizational chart (exhibit 1 in the case). The overall goal is to improve value for customers so perhaps it is better to emphasise and put more weight on departments who have immediate contact with the customers such as the 'customer problem resolution' department. Since one department affects another, perhaps it is also more important to improve the departments which have the most difficulties seen through measurements. There is a chance of biased data if employees have a close or long term connection to the person they are evaluating, even through checklist. It is important that data is as realistic as possible. Only through explaining to employees why this is important will allow for pragmatic results.
Improving the application form process
Application forms can be checked by one of the 80 branches before they are sent to VEG in Hamburg. Customers will fill in the most important sections in the application forms if the titles of the most important sections in the form are highlighted or placed at the top of the first page. This can now be all done online, where customers who can just fill in an online application form and an error box pops up if the customers have not filled in a required box, letting them fill it in again. This would save a lot of trouble in the whole application process, although this would still need to be checked by an employee. Since 49% of customers were outside Germany it would be a good idea to have application forms already translated in their own language so that customers can fill in the form easier.
Improving the employee's data processing
It must be measured to see if there is a large difference in inaccuracies between permanent staff and temporary staff. This also applies if there are there more inaccuracies during the morning shift or the evening shift (perhaps people are more tired or more relaxed). Perhaps there is a correlation depending on the number of working hours and number of errors. This should be measured, evaluated and dealt with accordingly. Employees can be offered incentives where the employees with the fewer errors in data are rewarded; this can create healthy competition and motivate employees. It is important that workers don't feel their job is threatened but it is also important for them to be liable to their mistakes. If certain employees are too tired when inputting data perhaps they would like to be asked whether they prefer the day or the evening shift. When an error has occurred to keep customers relatively satisfied, customers could be provided with a discount on their existing or next insurance period
Six Sigma Improvement Model
It would be beneficial for DAV to hire a Six Sigma professional who is an expert in projects that involve quality enhancement.
The six sigma model focuses on customer needs and process arrangement.
- Define: Customer service needs enhancement. Our aim is to decrease input errors that cause customers to be unhappy, while continuing with our fast pace workload.
- Measure: It is necessary to measure the number of errors as they have a direct influence on customer frustration. It is important to see at what point these errors are acceptable or not.
- Analyze: Find out what causes employees to input errors and rectify it. It is important to identify the specific variable (such as working long hours or working late). A fish bone diagram can help DAV see the breakdown.
- Improve: Here staff can be asked what they think can be improved as many factors are usually involved.
- Control: This can be done through the p-chart. Over time, the p-chart might give clues to where the process is heading, and whether it is predictable or not.
The p-chart (using data from exhibit 4 in the case)
total people in samples: 30x300=9000, total errors: 516
This means the average correct number of applications was 94.3% the p-chart is a measure of quality and should be used on an on-going process.
It is crucial to measure the lower and upper limits. If the measurements exceed the lower or upper limits there is a problem that should be rectified. Irregular fluctuations on the chart can give the company insight on opportunities to capitalise on, factors it should deal with, or threats that it should terminate.
As the p-chart is over time DAV can also see what time during the day employees are more likely to make errors (early morning or in the evening for instance). In this case adjusting when they work is an option. Full and part time workers can be compared and the statistics would show who is more productive. Experiments can also happen whereby the company can give certain employees more breaks and see their performance compared to other employees who feel more pressured; this would be to see if having breaks would lead to an increase in performance. Senior managers must understand that it would take some time to see results, as information, testing, altering the process and assessing it, is no quick task. It should pay off in the long run.
However this does not take into account the fact customers not only want a good insurance package; they also want polite employees. Apart from monitoring calls, customers can be encouraged to give feedback, employees can be given seminars on how to improve customer service, and even focus groups among customers and employees about how to improve the customer service can be introduced. Employees must be correctly guided and rewarded in order to perform successfully.