How to apply Six Sigma DMAIC for Reducing Client's Line Wait Time at the Bank Branch

Written by:

WHITEPAPER

Six Sigma DMAIC for

Reducing Line Wait Time

of Branch Clients

Jeremy Jay V. Lim Six Sigma Master Black Belt

Project Management Professional

D i s c l a i m e r : T h e o p i n i o n s a n d i n s i g h t s e x p r e s s e d i n t h i s w h i t e p a p e r i s s o l e l y f r o m t h e a u t h o r a n d d o e s n o t r e p r e s e n t a n y b a n k . T h e d a t a u s e d i n t h i s w h i t e p a p e r w e r e s i m u l a t e d t o r e c r e a t e n e a r a c t u a l e x p e r i e n c e i n a b a n k b r a n c h a n d f o r b e t t e r i l l u s t r a t i o n o f t h e t o p i c .

Whitepaper: Using Six Sigma DMAIC for Reducing Branch Clients’ Line Wait Time 2

Introduction

Waiting in line is one of the common causes of client complaints in the bank's branch outlet especially if it takes more than 30 minutes before a client reaches his/her turn at the counter. This problem is usually addressed by “just do it” method. However with this method the banks are still far from reaching a long-term solution that will significantly reduce incidences of complaints due to long line wait.

“Just Do It” is one of the long existing process improvement techniques where data analysis and

planning is not needed because one would only need to act and apply the instantaneous solution right away. Hence, the word “Just Do It” came about. With the “Just Do It” method, the results are only confirmed through trial and error experiments that take up time before one is able to address the problem.

This whitepaper intends to show the merits of using Six Sigma DMAIC method to reduce the clients’ waiting time at the branch. Simulated data were used for illustration purposes. A comparison of 2 process improvement methods were presented to show the difference in the outcomes if one bank were to use the “Just Do It” method versus the Six Sigma DMAIC method. The statistical software named SAS-JMP version 10 was used in the creation of the analytical graphs and statistical analysis.

Problem Background

Bank A has a branch located in a busy area in Manila city where the mix of transactions is as shown in the pie chart below. Majority of the transactions are Check deposit (22%), check encashment (15%), bills payment (17%), cash deposit (18%) and cash withdrawal (17%). The daily total transaction averages to about 510 transactions per day which are being processed by 2 tellers - 1 new hired and 1 experienced tenured teller.


Its branch is a typical small branch with 2 tellers, 1 new accounts officer, 1 branch officer and 1 branch manager. It has the floor layout as shown in the picture below.

Bank A branch has clients who complained of frequently waiting in line for more than 30 minutes. Bank A believes that this should be addressed right away with quick and obvious solutions without the need for time-consuming data gathering and analysis. Their goal is to reduce or eliminate incidences of customer complaints due to long line wait time for the branch clients.

Bank A just wants to solve the problem right away with common sense but felt that they were not getting close to their goal of reducing the client’s line wait time. In this whitepaper, we will illustrate how Six Sigma DMAIC can be applied to approach this problem differently to be able to get on to the right track towards solving this problem.


Case Study Bank A – “Just Do It” Approach Bank A implemented a number of process improvements which addressed some of the small issues but they felt that they are still far from achieving a long-term solution. They also decided in the beginning that there is no need to get relevant data to compare the baseline and the outcome, thus, they were not able to assess the effectiveness of their implemented process improvements.

Problem: Processing stops when a teller has to wait when the other teller is printing on the shared passbook printer Solution: Allocate 1 passbook printer for each teller Result: Eliminated instances of interruption or waiting for the shared passbook printer to be available. However, there seems to be minimal effect on client’s line wait.

Problem: There were some occasions when the 2 tellers are both processing clients with multiple transactions since the 2 clients came from a single line where all the clients have to queue up. Solution: Create 2 queue lines with 1 line for each teller. One teller (New Hire) will be for Single transaction clients only, and the other teller (Tenured) will handle both Multiple and Single Transactions Result: Reduced the complaints but some clients still complain of a long line wait time in the single transaction lane. Effectiveness cannot be assessed without supporting data.

Problem: 1st come 1st served. Honesty basis. Everyone has to stand in the line before a clients gets their turn at the counter for transaction processing Solution: Guard to give Queue number when client enters Result: Somehow made the clients less irritable since they can sit on the chairs while waiting for their number. But still some complaints are received due to long wait.

Problem: Even the VIP clients complain due to long line wait time Solution: Allow VIP clients to set appointment with the branch officer via phone. Teller will process their transaction during off peak times Result: This was successful since the VIPs were scheduled either during off peak time or when there is less client coming in. But there were more number of complaints received during peak times

Problem: Only 2 tellers doing OTC transaction when queue becomes longer Solution: New Account Officer (NAO) also process OTC transaction during peak time periods after being given authorization to perform teller function Result: This somewhat helped shorten the queue but only during the available time of the New Account Officer. The New Account Officer can only assist when there are no new clients coming in. Based on the branch's experience, the NAO is usually available 2PM onwards which coincides with the off-peak time of OTC transactions.

Problem: Only 2 branch tellers doing OTC transaction even when queue becomes longer Solution: Branch manager calls on the reserved pool’s teller when queue becomes longer Result: In most cases, the branch manager is not able to monitor the branch queuing and is not able to call in additional teller on a timely manner. There must be a trend that can help guide the branch manager in deciding on when to have additional teller to assist the 2 tellers.


Recommendation Bank A – Six Sigma DMAIC Method Six Sigma is a type of process improvement methodology while DMAIC are the steps. “D” stands for Define the problem and goal, “M” stands for Measure the baseline process before it is improved, “A” is for Analyze the possible root causes, “I” is for Improve the process with the identified, “C” is for Control and sustain the improved process so that the problem will not go out of control in the future. In this paper, recommendations on how Bank A can do the process improvement differently following the Six Sigma DMAIC will be presented.

Step 1: DEFINE Phase

Project scoping. Project scoping is one important step during the start of a project. Bank A should decide to focus their attention on resolving the client’s line wait and not treat it just like an issue of the day. The idea is to create a success story and then replicate the solutions to other similar branches in order to maximize the benefits out of this project. Without proper project scoping, there is a tendency to solve every concern even if it does not have impact on reducing the client’s line wait.

Define the problem and the goal. Afterwards, the project team may look at the customer complaint logs such as the branch logbook to confirm if there were indeed complaints related to long line wait times especially if it was beyond 30 minutes. Afterwards, the project team can hire contractual employees to help measure the line wait times of clients in the branch on a usual normal banking day.

Using the simulated data, the problem and goal may be stated as follows:

Baseline: 273 incidences with line wait of more than 30 minutes. Goal: Reduce the long line wait incidences from 273 down to 100 incidences

We should take it that this is going to be the first attempt to address this problem and that there will be a series of continuous process improvement initiatives until the complaints due to long line wait time is reduced to manageable levels.

Step 2: MEASURE Phase In the Measure Phase, the project team reviewed the process, looked in to the trend of clients' line wait time and identified what data or information they would need for analysis purpose.

Process Mapping. In order to understand the details of the problem, they looked at the process flow from the time that the client entered the branch up to the time that the client left the counter after their transaction was completed. With process mapping, the project team was able to identify the problem areas and the potential causes of having client’s line wait of more than 30 minutes. Please see the diagram of a Process Map below.


<Process Map diagram>

Queuing and Line Wait. The Queuing Concept can also be considered. It states that the line wait time is driven by the number of client arrival per hour and by the processing time per transaction. Given, Client Arrival per hour = 120 clients per hour Ave OTC Client Servicing Time = ave. 3.5 mins/client (is equivalent to) Transactions per hour = 60mins/hour / 3.5 mins/client = 17 clients per hour

The Line Wait Time’s computation formula is as follows: Line Wait time = [(Client Arrival per hour) / (Clients serviced per hour X Number of Tellers)] X Ave. Client Servicing Time Example: = [(120 clients per hour) / (17 clients serviced per hour X 2 tellers)] X 3.5 mins/client = 4 clients on queue at any time X 3.5 mins/client Line Wait Time = 14 minutes on the average

Line Wait can also be computed with these suggested tools and techniques

• 1 start-stop timer app at the counter to measure the OTC processing time • 1 clicker counter with the guard or contractual employee to count number of clients falling in line

by the hour after filling out the bank transaction request form

*Note that 1 client can have more than 1 transaction, but in the queuing concept we use the number of clients serviced instead of transaction counts.

Trend Analysis. Trend analysis should also be done as it helps in confirming how frequent that the long line wait time happens and possibly during what situation. In the simulated trend graph, we can see that the long line wait happens during peak times of the day.


<Trend graph>

Data Collection Plan. Just collecting available data is not enough for identifying and analyzing the root causes, so a data collection plan has to be created. All the possible and related causes to the client’s line wait time should be listed. The data that will help validate the effect of these causes should be identified and recorded in a data collection plan. See the sample of a Data Collection plan below.

Data needed

Definition

How to measure Data source

Line wait time

From the time that the client fell in line after filling out the bank transaction form up to the time the teller starts servicing the client over the counter

The time when the client fell in line should be written by the contractual employee or branch assistant on the time study coupon

Time written on the time study coupon

Client servicing time

From the time when the teller starts servicing the client up to the end of servicing

When the client reaches their turn at the counter, the teller should write the start time and the time when the servicing ends regardless of the number of transactions

Time study coupon

Teller type

Whether New Hired or Tenured experienced teller (more than 6 months)

Ask for the name and if the teller is New Hired or Tenured Experienced teller

interview

Transaction type(s) per client

Type of the client’s transaction inside the branch which is written in terms of code words in the time study coupon

Contractual employee or branch assistant may ask for the transaction type of the client and write the Transaction type code on the time study coupon. Teller may edit the transaction type code if necessary

Time study coupon

With this data collection plan, the supporting data that will be collected may look like the data as shown below.


which drive the clients’ line wait time. Please see the conceptualized Fishbone diagram below.

<screenshot of supporting data following the data collection plan>

Step 3: ANALYZE Phase

Identify Root Causes. The Fishbone diagram is used for laying out the identified possible causes and their relationships with each other. The identified causes can be grouped together according to major causes so it will be easier to think of a common solution. The Fishbone diagram is helpful in identifying which of the causes directly or indirectly contributes to the problem. In this case, the identified major causes are the transaction process work interruptions, client arrival trend, teller’s processing capacity and the processing time per transaction are the major groups of causes,

<Fishbone Diagram>

Validation using Statistical Tests. Using the data collected following the data collection plan, statistical test can be done to arrive with validated significant causes contributing to the long line wait time. SAS JMP v10 was used for the statistical tests and for creating the statistical graphs called Box Plot. The blue line in the box plot represents connection between the averages of the items being compared. The results are as follows:

a. Tenured experienced teller has a faster processing time than a New Hired teller in most transaction. Thus, this teller would have a higher capacity to process more clients than the new hired teller.


b. Passbook printer sharing or the act of waiting while the other teller is using the passbook printer prolongs the OTC processing time for both the New Hired and Tenured teller. However, the number of incidences of passbook printer sharing for this branch under study is not that frequent enough to affect the length of line wait. Based on the data, it only occurred 10 times during the 1 banking day that was observed.

c. Manual override by a Branch officer for processing transactions with high amounts adds to the

processing time. But this also only occurred 11 times during the observed day so this relatively has small impact.

d. Certain transaction types significantly take longer time than the others in terms of OTC

Processing time such as Deposit – Cash / Check, Cash withdrawal, Bills payment Check encashment. These transactions also have the highest transaction volume thus they have a large influence on the client’s line waiting time. These are the transactions whose processing flows should be streamlined so that the OTC processing time could be reduced further.


Volume of Transactions in 1 Banking Day Deposit-Check 110 Deposit-Cash 93 Withdrawal-Cash 87 Bills Payment 86 Check Encashment-same branch 77 Check Encashment-interbranch 15 Check Encash (same branch) +Override 11 Deposit-Cash & Printer Sharing 10 Withdrawal-Cash + Over ride 8 Accomodated Transaction 7

Withdrawal-Cash & Printer Sharing 6 Grand Total 510

These are the confirmed significant root causes that should be the focus of improvements a.) Processing time of New hire teller is slower than Tenured b.) Need to focus further reduction on the processing time of the transaction types with high volume

ie. Deposit Check/Cash, Cash Withdrawals, Bills Payment, Check Encashment – same branch, interbanch

Step 4: IMPROVE Phase

After identifying the root causes, we focus now our efforts in addressing the root causes rather than tackling numerous branch concerns at the same time where some may not actually be related to the client’s line wait time.

Significant root causes

a.) Processing time of New hire teller is slower than the Tenured teller b.) Long processing time of the high-volume transaction types ie. Deposit Check/Cash, Cash

Withdrawals, Bills Payment, Check Encashment – same branch, interbranch Quick Wins. Implement the same quick win items from the “Just Do It” but choose the ones which are considered to be drivers of the Client’s line wait time. Fishbone diagram is the recommended basis for choosing a relevant quick win item to implement.

Long-term solutions.

a.) Tenured teller should coach the New hire teller on the best practice techniques to increase productivity. Branch manager should track the productivity level of the new hire teller and incorporate it as part of the performance appraisal requirements.

b.) To address Cash-based transactions - Installation of a desktop cash counter to help sort the bills according to denomination. The teller should also sort the bills during lean times of the day.

c.) To address Bills Payment - The teller should suggest to the repeat clients that they may use the bank’s online bills payment facility. The bank should generally push forward online banking in order to ease up the client traffic at the branch.

d.) To address Check encashment – some improvements may be like increasing the amount threshold of check encashment so that the branch officer will not have to do manual over ride frequently. Another helpful improvement is to enhance the system so that the transacting branch


may have an access or view of the specimen signature of the check owner/holder so there will be no delay during interbranch confirmation.

In the Improve phase, as the improvement solutions are implemented for at least 1 month, the queue lengths and line wait should also be monitored to assess its effectiveness.

Step 5: CONTROL Phase

Six Sigma’s way of thinking teaches us that it is more effective if not only the complaints are monitored also its root causes. Six Sigma teaches us that to control the Y we should control the X’s. In other words, to control the outcome we should manage the inputs. After implementing the process improvement solutions, it is time to setup the following process controls and means to monitor the following items and plot it in a trend graph dashboard to show the status. Then, there must be a deliberate effort to sustain the improved branch performance after completing the Six Sigma project.

• Control item 1: The queue length should not be more than 10 clients. That would likely result to client line wait of more than 30 minutes. Countermeasure Plan: a.) Teller speed up on the processing, b.) Branch manager call in reserved pool teller or New Account Officer to assist

• Monitor and Create Trend Graphs for the following items:

a. Teller’s Average processing time and the gap between New Hire and Tenured Teller – based on the evaluation test every 6 months

b. Client Arrival Totals Daily and Hourly – based on the counter recorded by the guard c. No of Clients served per hour – time study coupon encoded in Excel by branch assistant d. Hourly Queue length and Line Wait average which is computed based on Client arrival per

hour and Number of Clients served per hour e. Monthly reporting of the average transaction processing time for the transaction types such

as Deposit Check/Cash, Cash Withdrawals, Bills Payment, Check Encashment – same branch, interbanch. The expected result should show trends of shorter processing time or a down trend.

Comparison

Prioritization. "Just Do It": Prioritization depends on the burning issue at the time. "Six Sigma DMAIC": Prioritization is based on what would have the biggest impact on solving the problem where data and knowledge of the process are success factors.

Root Cause Analysis. "Just Do It": There is an inclination to tackle the causes as issues independent from the problem itself. "Six Sigma DMAIC": Statistical tests using sampled data were done in a controlled environment to assess the impact of the identified individual and multiple causes to the main problem of client's line wait. This results to the "Funneling Effect" where one is able to narrow down the contributing causes from the trivial many,- down to most significant few causes.

Sustaining the Improvements. "Just Do It": There is a likelihood of monitoring and controlling of the outcome but not the root causes that drive the outcome. "Six Sigma DMAIC": Trend graph is used to track the status of the outcome or the Process Y which is the clients' line wait time and also tracking of the root causes or Process X's which drive the Process Y. In Six Sigma DMAIC it is proven that to reduce the clients’ waiting time, the bank has to focus on addressing its significant causes such as the OTC processing time and hourly client arrival trend as described in the Analyze phase.


Summary

Businesses usually preferred to apply the quick “Just Do It” method in solving problems. However, the use of Six Sigma DMAIC method also has numerous advantages. With the “Just Do It” method, time is lost as businesses do continuous trial and error experiments just to arrive with an effective solution. While in the Six Sigma DMAIC method, time is saved as the long-term solutions are applied on the validated root causes. Process improvements are also sustained through monitoring and control. With this method, we are able to take the right path and arrive at the destination earlier.

How to apply Six Sigma DMAIC for Reducing Client's Line Wait Time at the Bank Branch

Career