THE UNIVERSITY OF NEW SOUTH WALES School of Mechanical and Manufacturing Engineering Semester 1, 2013
GSOE9810 – Process and Product Quality in Engineering
Assignment 1: Case Study Manzana Insurance: Fruitvale Frui tvale Branch (Abridged) (Abridged )
Group Number: 13 Team Members: Last Name AKOMEAH BASSAN HE KASIM OROMIEHIE ZHANG
First Name Matilda Marco Jie Michael Ebrahim Yan
Lecturer In-charge: Dr Eric van Voorthuysen Tutor In-charge: Mr James Baulch
Student ID 3447443 3451903 3469345 3285455 3314896 3352792
1.
Process Flow Diagram
The Process Flow Diagram, or PFD, of Manzana’s Fruitvale branch is shown in figure 1. Some important features of the PFD are as follow: •
The coloured arrow is used in order to differentiate the origin of insurance request. As shown in the diagram, the RUN, RAIN and RAP are originated from the customer. On the other hand, the RERUN is originated from the branch record.
•
The black arrow explains the path that each policy undergoes before being delivered to branch record, customer and agent.
•
Automatic Renewal System represents the current computerised system to produce RERUN request.
•
The Branch Record is assumed to store both hard copy and soft copy of the policy, and the data is used as the input for Automatic Renewal System.
The elements involved in the PFD are used base on the general convention of process flow diagram, which is discussed by Foster, 2013.
2
Figure 1 PFD of Fruitvale branch
3
2.
Performance of Manzana’s Fruitvale Branch
The assessment of the branch performance is conduced based on two key measures. These two key measures are financial and operational performance.
2.1.
Financial Performance
The data used in analysing the financial performance is taken from the Financial Summary, Fruitvale Branch in Exhibit 5 of the case study. The key concept in
analysing this performance is based on the ability of the branch to achieve satisfactory results or positive trend on some important economic data. The failure of meeting these criteria indicates the existence of problems, or issues, that can be verified based on the operational performance. 2.1.1. Branch Profit
The profit is an accounting measure, and the value is obtained by subtracting revenue with expenses (Sullivan, 2012). In general, the revenue of a healthy company is more than the expenses, and therefore profit is generated, otherwise there will be financial losses.
2.
Performance of Manzana’s Fruitvale Branch
The assessment of the branch performance is conduced based on two key measures. These two key measures are financial and operational performance.
2.1.
Financial Performance
The data used in analysing the financial performance is taken from the Financial Summary, Fruitvale Branch in Exhibit 5 of the case study. The key concept in
analysing this performance is based on the ability of the branch to achieve satisfactory results or positive trend on some important economic data. The failure of meeting these criteria indicates the existence of problems, or issues, that can be verified based on the operational performance. 2.1.1. Branch Profit
The profit is an accounting measure, and the value is obtained by subtracting revenue with expenses (Sullivan, 2012). In general, the revenue of a healthy company is more than the expenses, and therefore profit is generated, otherwise there will be financial losses.
=
− … [1]
The profit of the branch is shown in the following figure.
BRANCHPROFIT 2500
y = -244.21x + 2257.7
) ( 2000 d n1500 a s u o h1000 T n i 500 t i $ o r 0 P
R = 0.94586 ²
-500 Q1
Q2
Q3
Q4
Q1
Q2
1989
Q3 1990
Q4
Q1
Q2 1991
Year
Figure 2 Branch profit
4
Some key points based on figure 2 are as follow. •
The trend of the profit is assumed to follow linear regression.
•
The reliability of the trend line is assessed based on the R-Squared value. A reliable trendline commonly has R-Squared value of 1 or near 1 (Microsoft, 2013). The R-Squared value of the profit trendline is 0.94586, thus the trendline is reliable.
•
The slope of the trendline is -244.21, thus the profit has a negative trend.
•
There are two significant financial losses in the last two quarters.
The negative trend and losses are mainly due to the gradual drop of revenue and not because of increase in operating expenses. The operating expenses increased by approximately 16.31% from the first quarter of 1989 to the first quarter of 1991, mainly due to Finance Charge and introduction Salary/Plus program. This increase is less significant in compare to the significant drop of gross underwriting result of approximately 56%. The computation is summarised in the following table.
Financial
Q1 1989
Q1 1991
Difference
Increase or
Data
(A)
(B)
(B-A)
Decrease (%)
1238
1443
205
16.56%
3003
1322
-1681
-55.98%
Operating Expenses Gross Underwriting Result
Table 1 Revenue and expenses data
Furthermore, the significant elements that construct the gross underwriting result are the net underwriting revenue and the ordinary losses. As can be seen in figure 3 below, there is minimum variation of net underwriting revenue throughout the period, while the ordinary insured losses has been drastically grown over the period.
5
GROSSUNDERWRITINGRESULTS ANALYSIS ) 10000 d n a 8000 s u o h 6000 T n i ( 4000 e u l a 2000 V 0 Q1
Q2
Q3
Q4
Q1
Q2
1989
Q3
Q4
1990
Q1
Q2 1991
Year GrossUnderwritingResults
OrdinaryInsuredLosses
NetUnderwritingRevenue
Figure 3 Gross underwriting results analysis 2.1.2. Market Share
The drastic growth of ordinary insured loss shows contradiction between market share and amount paid to the customer in term of ordinary insured loss. The amount of ordinary insured loss is commonly increased if the market share is increased as well, or there is a positive correlation. In the case of Fruitvale branch, the ordinary insurance loss increased, but the market share decreased due to high renewal loss. This issue can be addressed by analysing the ratio between ordinary insured loss and the total policy. The analysis is summarised in the following figure.
OrdinaryLosses/TotalPolicies 10.00 8.00 o i t 6.00 a R 4.00 2.00 0.00 Q1
Q2
Q3
Q4
Q1
Q2
1989
Q3
1990
Q4
Q1
Q2
1991
Year
Figure 4 Ratio between ordinary insured losses and total policies
6
The increasing ratio in figure 4 indicates the existence of problem in regard to the quality of the customer, performance of the agent and the company’s underwriting standard. These issues will be addressed in the later section of this assignment. 2.1.2. Renewal Loss
The lack of ability of the company to maintain their existing customer is represented by the high renewal loss. The renewal loss has a significant impact toward the profitability of the branch. The revenue generated by the renewals is shown in the figure 5 below. The revenue generated by the renewal starts to decrease from the first quarter 1990, and it cancels out the revenue generated by the new policies. Thus, it has a direct effect toward the profitability of the branch.
Renewal ) 7300 d n7100 a s u6900 o h t n6700 i ( e6500 u l a V6300 6100 Q1
Q2
Q3
Q4
Q1
Q2
1989
Q3
1990
Q4
Q1
Q2
1991
Year
Figure 5 Renewal (Dollar Value)
2.2.
Operational Performance
The operational performance of the branch is assessed based on three main measures. These measures include customer satisfaction level, employee attitudes toward company’s strategy and day-to-day operation performance. 2.2.1. Customer Satisfaction
The performance measures in regard to customer satisfaction in this case study are the turnaround time, late number and renewal loss rate.
7
Turnaround Time (TAT)
As indicated in the case study, the TAT has increased from three days to more than five days. This is a significant competitive disadvantage for Fruitvale branch knowing that agents typically send their clients to the company with the shortest TAT, mainly due to the opportunity for obtaining commission earlier or selling more policies. It is clear that in the insurance industry the TAT is a key item. The situation is worsens due to the fact that Golden Gate, Manzana’s main competitors, has decided to implement a guaranteed TAT of one day.
Thus, a significant change toward
improvement of TAT is necessary if Manzana want to stay competitive in the business. Late Number
Pareto Analysis is conducted in order to justify that the RERUN has the most significant affect toward the total late figure. The data for the analysis is taken from Exhibit 6 , and the result is shown in the figure 6 below. The result shows that RERUN
contributes to more than 80% of the total late figure. The high late figure on the RERUN indicates poor service toward the current customer, and thus it lead to poor customer satisfaction level.
ParetoAnalysisonLatePolicy 3000
100.00
e2500 t a L f 2000 o r 1500 e b m1000 u N 500
99.90 99.80 99.70
0
) % ( e g a t n e c r e P
99.60 RERUN
RAIN
RUN
RAP
TypeofPolicy Frequency
Cummulative Frequency
Figure 6 Pareto analysis in late policy
Furthermore, the reason behind the high renewal late is due to the priority that is given to new policy. The summary in Exhibit 6 shows that the late figure of RUN and RAP is zero, while RAIN and RERUN increase throughout the year. This issue is strongly related to the profit generated by new policy and the employee attitude
8
toward the implementation of Salary/Plus system, which will be discussed in detail in the later section of this assignment. Renewal Loss Rate
The high number of renewal late leads to high renewal loss rate. As shown in the case study, the renewal loss rate increased from 33% in the second quarter of 1990 to 47% in the second quarter of 1991. On the other hand, the renewal loss rate of Golden Gate, Manzana’s main competitor, is 15%. Although renewals do not generate high profit for the company, losing some market share each year is a major problem.
The lost of market share can be measured quantitatively by analysing the renewal rate figure, which is derived from the data given in Exhibit 7 . In 1989, renewal rate was 85.9% and it decreases to 69.2% in 1991. The computation is summarised in the following table. Request Summary
1989
1990
1991 (6 months)
RERUNs
5166
4987
2081
Renewal Lost
849
1717
926
Total
6015
6704
3007
Renewal Rate (%)
85.89%
74.39%
69.21%
Table 2 Summary of renewal rate 2.2.2. Inconsistency in Following Company Policy
The company has used first-in-first-out (FIFO) system as a policy for considering requests. However, this policy is not support by employees. As mentioned before, the employees tend to give priority to new policy. The discussion of the violation of company policy, which includes the cause and the affect, will be included in the next section of this assignment. 2.2.3. Uncontrolled Day-to-Day Operation
The data on policy processing time in Exhibit 4 shows that there is a lack of standardisation in performing a task. This issue is justified based on the high number of standard deviation, which mean there is a significant variation in performing a task. The worse case scenario is taken from the processing of RERUN in the underwriting department. The mean in performing the tasks is 18.7, while the standard deviation is
9
19.8. This issue will be further analysed in detail in the next section by using the statistical concepts known as coefficient of variation.
2.3.
Outcome
The outcome of this analysis is that the branch performance is poor in term of both financial and operational. In term of financial performance, the branch profit has a negative trend and there is a significant loss in the last two quarters. Next in term of operational performance, the ineffective day-to-day operation and poor attitude of employee toward company policy adversely affects the customer satisfaction level. The TAT has increased recently, and the renewal late and loss rate also increase recently which indicates poor customer satisfaction and lost of market share.
10
3.
Issue Analysis
This section focuses on the issues that have been addressed in the previous section as well as the causes of these issues. The following figure summarised the issue analysis of the Fruitvale branch.
Figure 7 Issue tree
11
3.1.
Relationship Among Problems
3.1.1. TAT and Renewals
The main relation of TAT toward renewal late and loss are mainly the use of TAT to set the due date of new policy. The new policy is currently prioritised in different department, thus the renewal process will be further delayed. This delay will increase the renewal late, in which turns out to be renewal loss. 3.1.2. TAT and Profit
The main relation of TAT and profit is that the agents use TAT as the main criteria to measure the branch’s capability. The high TAT leads to competitive disadvantages, thus it will be less attractive to the agents. Furthermore, it will lead to lower incoming customer, thus lowering the profitability of the company.
3.2.
The Causes of High TAT
3.2.1. Inconsistent Implementation of Operational Strategy
The original operational policy in each stage of the underwriting process is First-InFirst-Out (FIFO) system in which orders are processed in the same order in which they arrive (Mieghem, 2005) . The FIFO system in Manzana’s operational context can be defined as the documents, or requests, are processed in order of their arrival, regardless of the importance and the complexity of the requests. However, the implementation of this policy is considered to be ineffective. The ineffective implementation of FIFO system is caused by the lack of supervision, the shortage of leadership and authority as well as the implementation of “Salary/Plus” system.
The manger of each department shows a negative attitude toward correct implementation of FIFO policy. The manager in the underwriting department prefers to prioritise RUN and RAP due to its profitability, while the manager of the policy writing prefers to priorities simple task for the purpose of reducing the backlog. Furthermore, the lack of authority and leadership is shown in the attitude of manager of rating department. The manager recognise the issue and the affect of violating the FIFO policy, but the manager could not fix the problem due to the order given by the higher management.
12
The second cause of violation of FIFO policy is the company's strategy of giving incentive for each new policy completed and such incentive was not applied for the renewals. This quantitative performance measure leads to prioritisation of request, thus the FIFO policy is violated. Furthermore, the main outcome of this issue is the non-value added activity, which is sorting the document by priority. This activity leads to high backlog number, which adversely affect the computation of TAT. 3.2.2. Poor Resource Scheduling and Allocation
The poor resource scheduling is justified based on the unbalance workload throughout the period, and it results in either overtime or idle period. Next, the poor resource allocation can be analysed based on the utilisation rate in each department. The computation summary of the utilisation rate is shown in the following table. The utilisation rate shows that the rating and policy-writing department are overstaffed, which result in low utilisation rate.
Daily Activity Average Time Required
Distribution
Underwriting
Rating
Policy Writing
40 Requests
40 Requests
40Requests
40Requests
40min
30min
70min
55min 27.5hours(75% of
Total Time Needed
26.7hours
20hours
46.7hours
Capacity Available
30hrs
22.5hrs
60hrs
37.5hrs
Utilization rate
89%
89%
78%
73%
daily requests)
Table 3 Utilisation rate
The significant cause of this problem is the outdated data of processing time required. The inaccurate data leads inaccurate estimation of the resource and time required to complete the job. Further, the affect of poor resource allocation and scheduling is mainly due to the backlog that it creates. The large number of backlog affects the computation of TAT. 3.2.3. Unstandardized Working Procedure
The data provided in Exhibit 4 shows that the 95% SCTs are derived based on the policy processing time, and inconsistent processing time will lead to inaccurate SCTs.
13
The inconsistent processing time is justified based on the high standard deviation values in some of the process stages. The selected quantitative analysis in order to provide an objective measure of the inconsistency is coefficient of variation.
The coefficient of variation, or CV, is the relative dispersion, and it is computed as the percentage ratio of the standard deviation to the mean. The general interpretation of CV is that data with CV of 30% or more indicates lack of process control (Brown, 1998). The mathematical equation for CV is as below. Next, the CV for policy processing time in each stage is calculated using the CV equation, and the result is summarised in table 4 below.
%
=
×100%
Type of Policy
Coefficient of Variation
RUNs
RAPs
RAINs
RERUNs
Distribution
44.8%
49.8%
21.1%
22.1%
Underwriting
73.4%
64.5%
51.8%
105.9%
Rating
27.2%
21.0%
24.3%
12.8%
Policy Writing
14.5%
-
15.9%
19.0%
Table 4 Coefficient of Variation Result
The computation shows that the process variation in Distribution and Underwriting department is mostly out of control with respect of 30% limit as specified by Brown, 1998. The variation of rating department is relatively high, while the processing time in Policy Writing department is relatively in control.
Furthermore, in relation to the computation of TAT, the uncontrolled process leads to high variation and SCT, and therefore the TAT will be overestimated. The introduction of standardised procedure may be able to provide better process control, and therefore the TAT will be more accurate. The relation is shown in Figure 1 below.
14
Figure 8 Relation between process control and TAT
3.3.
The Causes of High Renewal Late and Losses
3.3.1. Unreasonable RERUN Processing and Due Date
The renewal losses and late rate are related to unreasonable RERUNS processing and due date. A computer automatically generated a RERUN 30 days before the anniversary of the policy and the anniversary became the due date. However, the RERUNS will be released at the last day before the due date which leads to the congestion of RERUNS and less flexibility of process, and this will ultimately cause the high rate of renewal late and losses and long time TAT.
3.4.
The Causes of Negative Profit
3.4.1. Poor Originating Agent Performance
The number of new policy written and the ordinary insured losses are the two main parameters that can be used to measure the performance of originating agent. Based on the financial summary in Exhibit 5, the net underwriting revenue is relatively stable while the number of ordinary insured losses increases rapidly from $4,176,000 in the first quarter of 1989 to $6,453,000 in the second quarter of 1991. The total number of policies is decreasing in this period so the rise of the ordinary insured losses can be explain with the selection of bad customers by the agent. In addition, the poor relationship and lack of communication between the branch and agents is one of the contributing factors of this issue. Thus, the agents might prefer other propertyinsured company due to lower TAT and result in a rapid increase of ordinary insured losses. 3.4.2. Implementation of Incorrect Strategy
The new management implement a new strategy in which the branch was divided into three territories based on the geographic position. The main purpose of this strategy is
15
to improve the responsiveness toward the agent. Nevertheless, from Exhibit 7 , it is clearly shown that from 1989 to 1991, the number of lost renewals increased from 849 to 1852 (two times of 926 in first 6 months in 1991) and the Gross premium was nearly constant during two years. As a result, it seems that the geographical restructuring is ineffectiveness and it has a significant impact to the company’s profitability.
16
4.
Problem in Turnaround Time (TAT) Calculation
The Turnaround Time, or TAT, is defined as the number of days between the receipt of a new commercial policy, or RUN, and the issuance of a final policy. The calculation of TAT by Fruitvale branch is included on Exhibit 3. The context of this calculation is based on the ending week of September 6, 1991.
There are some problems identified in regard to the current calculation of TAT. The following are a brief list of the identified problems: 1. Obsolete input data from Exhibit 4 2. Incorrect assumption of process start-up 3. Incorrect assumption of requests’ transfer procedure
4.1.
Obsolete Data
The TAT calculation in Exhibit 3 is taken from 1991 context, and it utilised the Standard Completion Time, or SCT that is constructed based on the policy processing time in 1986. These SCTs are considered to be obsolete due to the installation of computer system in Rating and Policy Writing Department during late 1980s and early 1990s. The installation of computer system has significantly reduced the processing time required to complete activities in these two departments. Therefore, the utilisation of the 95% SCTs of Rating and Policy Writing Department is not valid, and it leads to overestimation of TAT. Furthermore, the utilisation of the old data in the TAT calculation shows that the company does not have a clear strategy in managing their data.
4.2.
Inaccurate Assumption of Process Start-up
The computation of TAT in Exhibit 3 is conducted based on the assumption that, the process in the next department can only start once the process in the previous department has been completed, regardless of the requests that are currently in the queue. For example, the requests in the Underwriting Departments (3 RUNs, 7 RAPs, 6 RAINs and 36 RERUNs) cannot be processed until the requests in Distribution Department have been completed, or in other word the Underwriting Department needs to wait for approximately 0.6 days before they can start the process. The graphical representation of this issue is shown in figure 9 below.
17
Figure 9 Sequential process start-up assumption
The assumption of the sequential process start-up leads to overestimation of TAT. The alternative assumption that can be used to estimate TAT is by assuming processes in a different department start in a parallel manner. It means that each department will immediately process requests that are currently in the queue and does not wait for the previous department to complete their tasks. This assumption will reduce the estimation of TAT as well as improving its accuracy. The graphical representation of this assumption is shown in the following figure.
Figure 10 Parallel process start-up assumption
18
4.3.
Incorrect Assumption on Request’s Transfer Procedure
The current computation of TAT assumes that the requests are transferred in batch. This assumption is incorrect since the company does not implement a batch transfer policy, and it also results in overestimation of TAT. Thus, the individual transfer is preferable and the use of this assumption may improve the reliability of TAT. Furthermore, the utilisation of individual transfer required addition delay period to prevent underestimation of TAT.
19
5.
Recommendation to Fruitvale
The
availability
of
information
ranging
from
the
company’s
background,
organizational structure, operating procedures and the company’s performance is utilised for revising and modifying the company strategy and policy. This information serves as roadmap for the recommendation for further improvement.
The proposed recommendations for solving the identified issues are discussed in detail in this section. The discussions for each recommendation include the related issue, the recommendation proposal and its effect toward further improvement. Lastly, a matrix diagram is included as the basic guideline on the significance of recommendation toward further improvement.
5.1.
FIFO System and Penalty
Related issue: •
The teams in different department are currently prioritising the incoming request, thus the policy is violated.
Recommendations toward the issue: •
Emphasise the criticality of FIFO policy
•
Emphasise the importance of customer satisfaction, not the premium generated by the policy
•
Introduce supervision system to ensure the policy is followed, and allocate penalty
Effects toward improvement: •
Reduces the non-value added activities, such as sorting
•
Reduces the renewal late and lost due to the elimination of prioritisation.
•
Reduces TAT
•
Improve the net income
20
5.2.
Revaluate Turnaround Time Computation
Related issue: •
The high turnaround time is a main competitive disadvantage for the branch.
Recommendations toward the issue: •
Revaluate the method in estimating TAT due to some issues in the computation that have been addressed previously. The additional information for correct TAT computation can be obtained by benchmarking the computation.
•
Revaluate the input data for estimating TAT.
The effect toward improvement: •
5.3.
Improvement of TAT in term of low TAT and its accuracy
Revaluate Salary/Plus System
Related issue: •
The introduction of Salary Plus system leads prioritisation of new policy since the incentive will be paid only for each new policy written about quota.
Recommendation toward the issue: •
The incentive will be paid for all policy written above quota. This action will increase the operating expenses, but the benefits, in term of improvement in revenue, will overcome the increase in expenses.
Effects toward improvement: •
Enhancing the FIFO policy
•
Reduces the renewal late rate
5.4.
Revise the Agent’s Commission
Related issue: •
The significant gap of renewal commission (7%) and new policy (25%) motivates the agent to focus on getting a new customer despite of the quality of the client.
Recommendation toward the issue: •
Decrease the gap between the commissions by increasing the commission of renewal policy.
21
Effect toward improvement: •
The good quality customer will reduce ordinary insured losses, thus improving the branch’s profitability.
5.5.
Revise the Terrorise Strategy
Related issue: •
The division of territory strategy is introduced to improve the responsiveness, but this strategy in fact increases the renewal lost, thus it is ineffective.
Recommendation toward the issue: •
Bringing the underwriting teams together in order to centralised the policy processing.
Effect toward improvement: •
Even and balance workload for employee
•
Reduce the TAT due to better resource utilisation
5.6.
Guideline for Computerised System
Related issue: •
The implementation of computerised system to ensure on time RERUN processing is not effective due to lack of guideline.
Recommendation toward the issue: •
Establish the clear guideline in regard to receiving and processing RERUNs that are generated by the computerised system
Effect toward improvement: •
5.7.
Prevent unnecessary delay of RERUN process, thus reducing the renewal late.
Data Management System
Related issue: •
The out-dated data used in the TAT calculation leads to inaccurate estimation, thus a proper data management system needs to be implemented for the purpose of regular data update.
Recommendation toward the issue: •
Implementation of data management system such that regular update from the main company can be obtained in a real time manner.
22
Effect toward improvement: •
Improve the accuracy and reliability of TAT calculation and other related computation.
5.8.
Standardise Working Procedure
Related Issue: •
The large variation of the process in each department leads to high TAT computation.
Recommendation toward the issue: •
Establishing a standard working procedure that provides clear guideline of the step-by-step procedures that need to be done for different policies in different department. The use of graphical representation, such as flowchart, can be implemented to improve the effectiveness.
Effect toward improvement: •
The low variation of process leads to lower Standard Completion Time (SCT), thus it reduces the overall TAT estimation.
5.9.
Outcome
The matrix diagram is constructed in order to show the priority of the recommendation and inform the management regarding the significance of each recommendation. The matrix diagram is shown in the following figure.
Figure 11 Matrix diagram
23
Bibliography Brown, C. E. (1998). Coefficient of Variation. Applied Multivariate Statistics in Geohydrology and Related Science, 1998, pp 155-157 Foster, S. T. (2013). Managing Quality: Integrating the Supply Chain (5th ed). Pearson Education Limited, Harlow
Microsoft Corp. (2013). Choosing the best trendline for your data. Retrieved on April 29, 2013 from http://office.microsoft.com/en-au/help/choosing-the-best-trendline-foryour-data-HP005262321.aspx
Mieghem, P.V. (2005). General Queuing Theory. Retrieved on April 29, 2013 from http://www.nas.its.tudelft.nl/people/Piet/CUPbookChapters/PACUP_queueing.pdf Sullivan, W.G., Wicks, E.M. & Koelling, C.P. (2012). Engineering Economy (15th ed). Pearson Higher Education, Upper Saddle River, New Jersey
24
