Managerial decision making (USA World Bank)

8/14/2019 Managerial decision making (USA World Bank)

1/57

USA World Bank

09Problem SolutionRubricMBA/510 MANAGERIAL DECISION MAKINGAlbee HorneProfessorErick Ahrens


2/57

USA World Bank (UWB) with domestic and international presence has been exploring

alternative ways to advance their market share. Their main area of focus for now is the launching

of Instant Reward Instrument and Instant Reward Data a reward card that is being

developed to get potential customers to switch to UWB. Throughout this paper I will be

explaining the concepts of valid and reliable research, analyzing descriptive statistics, exploring

different sampling methods, applying basic probability concepts related to normal distribution,

interval confidence levels in relation to making business decisions, and solving business

problems.

The above will be the tools used in helping to solve the problems that UWB will face in

initiating this project. Once the correct problem is framed, and research is validated, and the

correct statistical calculations are in place, end goals can be created. After the creation of the

end goals, I intend to create a step by step plan to meet those goals throughout this paper.

Lets begin with the key players who all will have a hand in making the launch a success.

Please take a look at the list of key players involved along with information about their area of

expertise, and how their main characteristics relate to the project.

Brian Allen, President of New Product Development for USA World Bank:Brian is a 13 year veteran who has had experience in the past with great product launches. Brianand his marketing team is responsible for developing a successful project.

Mary Monroe, Vice President of New Product Development: Mary is a10 year veteran with agood track record. The final choice of product has to be approved by Brian, her immediatesupervisor. Even though she may have a good track record, she is not thorough with her investigations concerning the validity of third party research.

Jim Wilson, Vice President of Marketing Development for USA World Bank: Jim has beenwith the company four years and has grown the small business segment 40% with his leadershipskills (no nonsense type demeanor), building a strong bond with small business owners. He also

Page2


3/57

feels a connection with, Aaron Anderson, Beth Brown, and Charlie Cousins, since they areclosest to the customers and their feedback.

Alexis Andrews, Chairman of the Board of USA World Bank: She has been with UWB for 3years, and believes in hiring those who fit the companys strategy. She is a firm believer of getting things done, without errors.

Bea Hansen, newest member of the Board of Directors at USA World Bank: She has servedon the board for three months, and has a strong back ground in statistics. Any decision that shemakes will be based on what can be proven, and what is factual.

Tom Araya, Marketing Associate: He recently earned his MBA and has some experience withstatistical analysis.

Utilize a Problem-Solving Approach

Before we can arrive at any end goals, one starts with defining the problem that is

plaguing UWB. One of the major questions is whether or not the data used to conclude the

survey was sufficient enough, for the basis of making the best decision regarding the next course

of action, regarding this major project. This would also lead to the question to whether the

statistical data was deciphered correctly, as well as if the correct markets were targeted, in order

to meet the companys end goals. We have Jims statement from the scenario quote Tell me

more about the focus groups. How were the participants selected? How were groups conducted?

Can I see the survey we used? These are good questions which act as fuel for defining the

problem in order to develop a probable solution.

In an attempt to try to seek answers to the above questions we start with the validity and

reliability of data.

Explain the concepts of validity and reliability of data

Page3


4/57

According to Cooper & Schindler (2003), quote Omission of significant procedural

details makes it difficult or impossible to estimate the validity and reliability of the data and

justifiably weakens the confidence of the reader in the research itself as well as anyrecommendations based on the research. This characteristic is comparable to developing a

tactical plan. This statement brings us back to Jims statement, requesting verification of sample

groups and participants, as well as requesting to see the actual questions used on the surveys,

clearly this is the recognition needed to confirm the reliability of the data. One way of doing that

is deciphering the correct sampling method

Sampling methods

When the committee had their meeting a few facts were revealed about the sampling

methods used to conduct the survey. Some of the questions that surfaced were the sample size

used, sample type (Random) and the focus group that was targeted. In addition to that

information another factor considered was the time frame in which all of these samples were

taken. In order to validify and determine reliability of the proposed data the following questions

were asked by Bea quote Excuse me, Brian, could you please give me an indication of how

large a sample was selected to get to this conclusion? Could you also tell me the time frames in

which this data was collected? (Scenario University of phoenix 2009)

According to Lind the best way to find out about a population is to get a sample size. The

next question, is what is a good sample size, and which way would be the best way for obtaining

one? A huge sample size may distort the results; the sample size was 140,000 which is a large

sample.

Analyze data using descriptive statisticsPage4


5/57

Upon Analyzing the data at hand Bea realized that the survey was conducted on line and

a majority of the participants were male and possibly of a younger age and possibly more

upscale. She also mentioned a few flaws within the survey, quote :

Youre showing a higher percentage of men in this sample, possibly because men are morecomfortable with computer technology and therefore more likely to respond to an online survey.I would be a lot more confident with a truly random sample. I mention age and income because

its well-known that tech-savvy people tend to be younger and more upscale. (Scenario P.8 )

The lesson here is that banking business and trends, like everything else in our world, changesquickly. If you can just compare the survey results from February and March of 2004 with thoseof January and February of 2005, Id feel a lot more secure about your data. (Scenario P.8,

paragraph 4)

Taking all of the above into consideration the overall concept is that when dealing with a

biased population, a sampling error can result, according to Lind (2004) the definition of a

sampling error is the difference between a sample statistic and its corresponding population

parameter. The equation is an example of how a sample is conceived while is

an example of how to find the sampling error.

Another type of sample which would be beneficial is a random sample, in which the

definition states that this type of sample is one where each item or person in the population has

the same chance of being included (Lind 2004). This will be possible only if the participants

were more random and less specialized or biased. A method other then internet surveys should

be used in order to utilize the random sampling method.

Apply basic probability concepts to facilitate business decision making

Page5


6/57

I see a direct correlation between age and dependency to buy. This can be represented by

a Linear Regression. This there was a statement made by Bea which also solidifies this fact.

A significant correlation, such as you found, means that the coefficient, or weight of the

mathematical model, is unlikely to equal zero. It means that there is a relationship between age

and the likelihood of purchase and the likelihood of changing banks. The information about age

helps us estimate purchase and bank changing. (Scenario pg 9 last paragraph)

Apply confidence intervals in solving business problems

The comment above references the level of confidence, which in most cases one would

hope to be 90% or better. In fact Bea also made a reference to this in the 4 th paragraph page 10 of

the Scenario

There is an old saying in my business that an R-squared between 70 and 80 percent is agood working model. R-squared between 80 and 90 percent is excellent. R-squared between 90and 95 percent qualifies the researcher for the statistician's hall of fame. And an R-squared above95 percent means that you are probably lying.

The higher confidence level means the higher the possibility of the desired results to

occur within the mean. Since we know 20% is extremely low, we can at least decide how much

higher that confidence level would need to be in order to predict the desired results.

For UWBs Instants reward card data, the data in reference to tables 1, 2 and 9, the

probability of occurrence of likelihood of purchase by number of benefits, likelihood of changing

banks by number of benefits, likelihood to purchase and change banks by age, all had a

probability of less than .05 which corresponds to less than a 95% confidence interval.

Page6


7/57

Apply the normal distribution to facilitate business decision making

The following distribution is classified as a normal probability distribution

(Lind 2004)

You also have the mean

The deviation

The significance here is that when you are able to find the deviation (difference of the

mean) this could help you better analyze the problem in order to make a business decision. For

UWBs Instants reward card data, the data used to populate the statistical probabilities were

derived from this basic formula. In conjunction with this elementary principle, we can use this

data to determine regression lines and degrees of freedom, better known as Inferential Statistics.

Apply inferential statistics in solving business problems

Page7


8/57

The F distribution was used to compare different means simultaneously for the different

rewards data card (IRCD) data sample. The F distribution is defined as ANOVA or analysis of

variation. The formula for such distribution is (Lind 2004) and these are the steps

involved when testing your hypothesis.

1. State the null and alternate hypothesis

2. Select the significance level

3. The test statistic will follow the F distribution

4. The critical value is obtained, and in two tailed test can be found by the formula

5. Select the sample and perform the test.

This is how the calculation starts referencing (IRD) survey, s 2/1= 38.753 while s 2/2= 14.300

So F= 2.71

Page8


9/57

(Lind 2004)

To begin (IRD) we calculate the mean by dividing the sum of observations by their

number or total observations. Next the deviations are found and squared for all observations.

This equals the SS total for the (IRD) which is 38.753. The SSE = the difference in the

treatment and error is the degree of freedom which is 1(table 3 IRD appendix 1). The degree of

freedom indicates the chance of committing a type one error, and rejecting the null hypothesis

below.

Page9


10/57

Hypothesis 2 Age is negatively correlated to IR3 Likelihood of Purchase Top Benefit Only.

Keep in mind that the end-state goal

In reviewing the scenario the end- state goals for this company is apparent and are listed

below.

1. Focus groups will be selected to represent the overall population

2. Separate market research will be necessary if there is a difference in the focus groups

3. UWB will launch an instant card reward program for consumers and small business

owners.

There is a substantiated need to also analyze the small business owners fit into the

equation in regards to rewards. According to Jims statement from the scenario, he sees the

potential success of the companys financial future. He has recognized the important factors that

would play a role in survey data, like the size of the company, and whether or not they are the

founders of the company.

Brian is focusing on the statistical points in relation to survey data for these groups which

leads us to our first end goal, quote

But even in focus groups, we have to watch the response rate. If only 25 percent of the

companies are willing to take part in the focus groups, Dr. Hansen might say that they somehow

are different from the ones who refused. I don't know what the response rate should be in focus

groups, but in survey research, the rule of thumb is that a 50 response rate is considered

adequate, 60 percent is considered good, and 70 percent is excellent. (Scenario pg. 120)

The statement below from Brian leads us to our second and third end goal quote

Page10


11/57

So we need to make sure that the companies the market research company interviews represent

a full range of sizes and a good mix of founders and CEOs who came into the business after it

had started.

The step by step process as mentioned earlier will follow the below format.

1. Small sample (smaller sample will prevent distortion) 75-80 people is a good sample amount

2. Focus groups should be divided into those representing the consumer and those representing

the small business owners. We cannot ignore one profitable facet over the other.

3. Research should be handled by specialized individuals Jim and Brian may be better suited to

handle small business research, the construction of the survey, while Tom Araya, may be needed

to interpret the data. It looks as if 75 to 80 customers surveyed was the proposed amount to be

stratified and randomly chosen which will represent either group. I believe that amount should be

chosen for each group for a better representation of the overall focus group without stratification,

after all we still have a biased group from the previous survey illustrating that most of those

surveyed were men in one account of one survey, we wouldnt want this to be the case for the

small business owners as well. In addition there needs to be a mixture of founders and CEOs to

for an accurate sample.

4. Different specialists and departments would need to be brought on board to help coordinate

efforts. This will pertain to Finance, IT, and accounting. These departments can aide in a system

to estimate ROI over a period of time. This will be a factor in making the best business decision.

5. When an adequate sample is taken, showing the difference squared of the treatments and a

decrease between the mean deviation, we can assume a higher confidence level which will

Page11


12/57


13/57


14/57

to Create Value for the Business. University of Phoenix 2009

Under problem solving materials tab. University Of Phoenix 2009

USA World Bank Problem Solution and Defense. University of Phoenix 2009

Appendix 1

(14-27)

Instant Reward Product Data

Appendix 2

(28-36)

Small business Card

Appendix 3

(pg 37-56)

SCENARIO: USA World Bank

Page14


15/57

Appendix 1

USA World Bank

Instant Reward Product Data

Descriptive Statistics

1 Demographics

B5 Marital Status

1 Married 65.0

2 Not married 35.0

D1 Gender

1 Female 49.3

2 Male 50.7

D2 Education1 Less than high school 16.2

2 High school grad 27.8

3 Less than 1 year college 7.4

4 More than 1 year college no degree 14.8

5 Associate Degree 7.3

6 Bachelors Degree17.1

7 Professional Degree 6.3

8 Masters Degree 2.1

9 Doctoral Degree 1.0

Page15


16/57

D3 Age

Mean 36.22

Standard deviation 22.69

D4 Hispanic

1 Yes 12.3

2 No 87.7

D5 Race

1 American Indian/Alaskan Native .92 Native Hawaiian/Pacific Islander .1

3 Black 11.5

4 Asian 3.9

5 White 76.6

6 Mixed 2.1

7 Other 4.9

D6 Dependants at home

None 61.4

One 15.5

Two 15.1

Three 5.0

Four 1.8More than four 1.2

D7 Household Income

1 Less than $20,000 29.6

Page16


17/57

2 - $20,000 to $30,000 11.9

3 - $30,000 to $40,000 11.1

4 - $40,000 to $50,000 9.6

5 - $50,000 to $60,000 8.1

6 - $60,000 to $70,000 6.6

7 - $70,000 to $80,000 5.0

8 - $80,000 to $90,000 3.8

9 - $90,000 to $100,000 2.8

10 - $100,000 to $110,000 2.2

11 - $110,000 to $120,000 1.612 - $120,000 to $130,000 1.3

13 - $130,000 and over 6.4

2 - Intent to Purchase

IR1 Likelihood With No Information

Mean 3.64

Standard deviation .82

n 140,630

IR2 Product Benefit Rank average rank

4 Credit card interest 4.41

3 Loan interest 4.26

6 Services discount 3.875 Product discount 3.46

1 Free checking 3.29

2 No minimum balance 2.91

7 Other 2.30

Page17


18/57

Likelihood to Sign Up

IR3 Top benefit

Mean 3.58


n 140,820

IR4 Top two benefits

Mean 4.16

Standard deviation .75n 140,444

IR5 Top three benefits

Mean 4.25


n 139,947

IR6 Change banks - top benefit only

Mean 2.96


n 140,611

IR7 Change banks - top two benefitsMean 3.24


n 140,488

Page18


19/57


20/57

Acme Bank

Checking 24.7

Savings 21.0

Credit card 30.6

Home or auto loan 7.8

Other loan 13.4

Other 2.5

100.0

Mutual Savings

Checking 33.3

Savings 29.5

Credit card 22.4


Other loan 2.2

Other 2.2

100.0

U.S. Trust

Checking 39.8

Savings 27.8

Credit card 18.2Home or auto loan 8.5

Other loan 4.0

Other 1.7

100.0

Page20


21/57


22/57

Other

Checking 24.3

Savings 29.0

Credit card 11.2


Other loan 15.0

Other 7.5

100.0

B1-1 through B1-8 Distribution by percent of accounts

Checking

MegaBank 18.7

Acme Bank 17.6

Mutual Savings 11.7

U.S. Trust 13.4

USA World Bank 16.8

Local/regional bank 11.9

Credit union 5.0

Other 5.0

100.0

SavingsMegaBank 17.5

Acme Bank 16.3

Mutual Savings 11.3

U.S. Trust 10.2

Page22


23/57

USA World Bank 16.7


Credit union 8.6

Other 6.5

100.0

Credit card

MegaBank 22.6

Acme Bank 23.8

Mutual Savings 8.6U.S. Trust 6.7

USA World Bank 19.2


Credit union 1.7

Other 2.5

100.0

Home or auto loan

MegaBank 13.4

Acme Bank 10.8

Mutual Savings 7.1

U.S. Trust 5.6

USA World Bank 11.2Local/regional bank 14.9

Credit union 32.0

Other 5.2

100.0

Page23


24/57

Other loan

MegaBank 20.1

Acme Bank 23.9

Mutual Savings 1.9

U.S. Trust 3.3

USA World Bank 18.2


Credit union 19.6

Other 7.7100.0

Other

MegaBank 23.5

Acme Bank 10.6

Mutual Savings 4.7

U.S. Trust 3.5

USA World Bank 16.5


Credit union 25.9

Other 9.4

100.0

B2 Changed Banks

Yes 21.6

No 78.4

Page24


25/57

B3 Accounts Changed

Checking 18.6

Savings 20.8

Credit card 26.2

Home/auto loan 11.3

Other loan 14.0

Other 9.1

100.0

Inferential Statistics

Hypothesis 1 The likelihood of purchase will increase with the number of favorite benefitsoffered.

Table 1

Likelihood of Purchase by Number of Benefits

Standard Lower Upper Variable Mean n Deviation Bound Bound

IR1LikelihoodNo Information 3.64 1 140,630 0.82 3.64 3.64

IR3Likelihood-Top Benefit 3.58 2 140,820 0.72 3.58 3.58

IR4-Likelihood-Top Two Benefits 4.16 3 140,444 0.75 4.16 4.16

IR5-Likelihood-Top three Benefits 4.25 139,947 0.68 4.25 4.25

1 IR1 lower than IR4 and IR5 p


26/57

Hypothesis 2 Age is negatively correlated to IR3 Likelihood of Purchase Top Benefit Only.

Table 2

Regression of Age on Likelihood of Purchase and Change Banks

Model R R-Squared

Age by IR3-Likelihood-Top Benefit Only -.112 .0125

Age by IR4-Likelihood-Top Two Benefits -.211 1 .0445

Age by IR5Likelihood-Top Three Benefits -.346 1 .1197

Age by IR6-Change-Top Benefit Only -.167 .0279Age by IR7-Change-Top Two Benefits -.203 1 .0412

Age by IR8-Change-Top Three Benefits -.426 1 .1815

1 Significantly different from 0; p


27/57

Table 4

ANOVA-Regression of Age on IR4-Likelihood of Purchase-Top Two Benefits

Source df SS MS F P

Model 1 36.171 36.171 4.01 .04

Error 140,442 1,266,786.840 9.020

Total 140,443 1,266,823.011

Table 5

ANOVA-Regression of Age on IR5-Likelihood of Purchase-Top Three BenefitsSource df SS MS F P

Model 1 178.149 178.149 8.21 .00

Error 139,945 3,036,666.555 21.699

Total 139,946 3,036,844.704

Table 6

ANOVA-Regression of Age on IR6-Change Banks-Top Benefit Only

Source df SS MS F P

Model 1 77.703 77.703 3.96 .05

Error 140,609 2,759,451.625 19.622

Total 140,610 2,759,529.328

Table 7

ANOVA-Regression of Age on IR7-Change Banks-Top Two Benefits

Source df SS MS F P

Page27


28/57

Model 1 81.164 81.164 5.04 .03

Error 140,486 2,262,386.544 16.104

Total 140,487 2,262,467.708

Table 8

ANOVA-Regression of Age on IR8-Change Banks-Top Three Benefits

Source df SS MS F P

Model 1 105.528 105.528 9.08 .00

Error 140,360 1,631,263.920 11.622Total 140,361 1,631,369.448

Hypothesis 3 The likelihood of changing banks will increase with the number of benefits.

Table 9

Likelihood of Changing Banks by Number of Benefits

Standard Lower Upper

Variable Mean n Deviation Bound Bound

IR6Change-Top Benefit 2.96 1 140,611 0.87 2.96 2.96

IR7-Change-Top Two Benefits 3.242

140,488 0.80 3.24 3.24IR8-Change-Top three Benefits 3.89 140,362 0.75 3.89 3.89

1 IR6 no different from IR7 and lower than IR8 p


29/57

Appendix 2

USA World Bank

Small Business Card Product Data

Descriptive Statistics

1 Demographics

D1 Years in Business

Mean 15.54 Standard Deviation 7.68

n 78

D2 Founder

Yes 61.5 %

No 38.5 %

D3 Entered Business ____ Years Ago

Mean 11.67

Standard Deviation 4.80

n 30

D4 When Became Boss?

Mean 6.87

Page29


30/57


n 30

D5 - Type of Business

Agriculture 9.0%

Mining25.6%

Manufacturing29.5%

Business Service 24.4%

Personal Service 11.5%

D6 Number of Current Employees

Mean 4.62

Standard Deviation .86

n 78

D7 Number of Employees Typical

Yes 75.6 %

No 24.4 %

D8 Usually Have

More employees 52.6 %

Fewer employees 47.4 %

D9 Number of More Employees

Mean 2.10

Page30


31/57


n 10

D10 Number of Fewer Employees

Mean 2.89


n 9

2 Banking Relationships

B1 Banks and AccountsMegaBank

Checking 14

Savings 4

Credit card 10

Line of credit 2

Equipment loan 3

Other loan 3

36

Acme Bank

Checking 10

Savings 3

Credit card 7Line of credit 5

Equipment loan 3

Other loan 1

29

Page31


32/57


33/57

Local/regional bank

Checking 2

Savings 0

Credit card 2

Line of credit 1

Equipment loan 0

Other loan 0

5

Other

Checking 0

Savings 0

Credit card 0

Line of credit 0

Equipment loan 1

Other loan 0

1

B2 Switched Banks

1 Yes 26.9 %

2 - No 73.1 %

B3/B4 Which Accounts

Checking

1 Convenience 1

2 Price 3

Page33


34/57

3 - Other

Credit card

1 Convenience 1

2 Price 1

3 Other 2

Line of Credit

1 Convenience 1

2 Price 83 Other 0

Other loan

1 Convenience 1

2 Price 1

3 Other 2

B5 Gross Annual Income

Mean $1,874,166

Standard Deviation 447,882

n 78

3 Small Business Card Data Percent of comments(n = 78, may find rounding error)

Transcripts of the actual comments will be sent to board members upon request.The comments will be grouped according to the categories below.

Audio tapes of the sessions are also available.

Page34


35/57

Comments were categorized by two independent raters. A third rater settleddisagreements.

SBC1 Without a lot of knowledge about the product, what do you think?

Good idea 38.5 %

Nothing special 24.4 %

Cant tell from information provided 21.8 %

Miscellaneous comments 15.4 %

SBC2 Would a lower rate of interest on loans be an attractive benefit for thosewho qualify?

Yes 51.3 %

No 9.0 %

Need to know how much lower 19.2 %

Lower rate traded off against what? 15.4 %


SBC3 Would higher rates of interest on savings and the like be an attractive benefit?

Yes 66.7 %

No 9.0 %

Need to know how much higher 3.8 %

Lower rate traded off against what? 19.2 %


SBC4 What other benefits would be attractive to you?

Lower/no cost banking 16.7 %

More personal attention 10.3 %

Page35


36/57


37/57


38/57


39/57

very collaborative, and knows how to work corporate politics to her advantage. If she has aweakness, its that she is not always thorough in her assessment of the research from firms thatshe hires. Each year, with Brians approval, Mary is responsible for deciding which new productto present to the Board. She and Jim Wilson, the Vice President of Marketing Development,dont always see eye-to-eye on this choice, but she respects him as a person and professional.

Jim Wilson, Vice President of Marketing Development for USA World Bank: Jim has beenwith the company four years and in the past three, hes increased USA World Banks small

business segment by 40 percent through successful relationship building with small-businessorganizations and owners. Jim thinks he is in sync with this business segment because heconnects well with relationship managers like Aaron Anderson, Beth Brown, and CharlieCousins who are a direct link to the customers. A no-nonsense leader, Jim is very focused on

building business for the bank.

Alexis Andrews, Chairman of the Board of USA World Bank: Before coming to USA WorldBank three years ago, Alexis was a successful entrepreneur, owning and managing a smallcommercial bank. She has a long history of launching successful banking products, and of hiring

people who have complemented her strengths, yet compensated for her weaknesses. She is aresults-focused leader who has little patience for errors.

Bea Hansen, newest member of the Board of Directors at USA World Bank: Bea has beenon the Board for three months. A statistics instructor at a local university, she likes to makedecisions based on facts. She is not political and will not play like a team member just to reach adecision. She prides herself on being reasonable, however, and will support the decisions made

by Alexis Andrews.

Tom Araya, Marketing Associate: An employee of Marys, Tom is usually called on to helphis boss interpret statistical data. He has recently earned his MBA.

JANUARY 17EXECUTIVE MEETING

MARY: Im sorry that Brian couldnt be here today, but he did want me to go ahead with this

Page39

d sate,

cialty


40/57

meeting because hes interested in your opinions. Given that, Ill brief you on this yearsrecommended product and what the study feedback has been.

This year, were proposing a credit card that will work like a frequent flier program, wherecustomers can earn rewards as a result of their purchases with the card. We plan to partner withan airline, some hotels and retailers. Were calling it Instant Rewards, and its been compared tosmall business credit card or a card with discounted interest rates.

Weve had Best Market Research conduct some feasibility studies for this product, and I havethe results of those studies here. (Refer to Instant Reward Instrument and Instant RewardData.) According to Best, this product will be successful in our market. In fact, their researchindicates that the card could actually move people to switch from their current bank, if theincentives are strong enough. Youre all in research, so you can see how this objective studysupports the introduction of the product.

Im excited about this because weve been behind the eight ball when it comes to consumer rewards, and weve lost customers to banks who offer them. I agree with Best that there is littlequestion about next year's new product launch. Are there any questions here?

(Mumbling around the table concerning previously disregarded suggestions for this type of product.)

MARY: Do we have your support for this new product?

(A unanimous yes.)

MARY: Good! That went much more quickly than I expected so, since I know how busy you allare, if there are no further comments or questions, our meeting is adjourned.

JANUARY 24

EXCERPT FROM SENIOR MONTHLY RELATIONSHIP MANAGER MEETINGHELD BY JIM WILSON

Page40


41/57

JIM: Okay, that concludes our regular business. Lets discuss our new product for small business owners. What is the status of the analysis on the reward card?

AARON ANDERSON: My accounts have told me theyre considering other banks that havequicker access to capital when they need it. If we dont get this product launched, Im afraid Illlose 40 percent of my customers.

JIM: 40 percent? What are you basing such a large loss on?

AARON: The large number of customers giving me feedback. Its not pretty.

JIM: Youve discussed these products with them?

AARON: Of course.

BETH BROWN: Ill venture that about 30 percent of my business owners have been waitingquite a while for something that will enable them to easily buy small capital products, and theylike the idea of earning rewards. So I guess Im hearing the same thing.

JIM: Okay, so most of you are hearing the same things. Thats good, but Id still like some datato back this up. Where are we on the analysis?

CHARLIE COUSINS: We each conducted focus groups in our regions, which showed anoverwhelming demand for the product. Several small business owners brought up a possibleconsumer advantage as well, but they want to see the card for small businesses first. When areyou presenting this idea?

JIM: Tell me more about the focus groups. How were the participants selected? How weregroups conducted? Can I see the survey we used?

CHARLIE: Yes, sir. It's right here. (Refer to Small Business Card Instrument.)

Page41

d Stateal

cusups willselected wa


42/57

JIM: Great! As soon as I have the final results in hand, I'll move forward. When will you havethose?

CHARLIE: Ill have it to you by the end of the week, and Ill also include our procedure for arranging the groups and selecting the participants.

JIM: Excellent. Thank you all; keep up the great work.

JANUARY 27

E-MAILFROM: Charlie Cousins

TO: Jim Wilson

RE: Results reportCC: USA Relationship Managers

Attached is the report we discussed in the Monday meeting. Please contact me with anyquestions or recommendations. (Refer to Small Business Card Data.)

FEBRUARY 3

DIALOGUE EXCERPT FROM EXECUTIVE STAFF MEETING HELD BY JIMWILSON

JIM: So what do you think? I hope youre as excited as I am by the outstanding work that myteam did in identifying an opportunity for our small business clients. Im interested in your concerns, questions, or suggestions about launching this product.

EXECUTIVE 1: What methodology was used to identify this need?

JIM: My managers used focus groups of small business owners within each of their respectiveregions. Among them, they cover the entire country, so Im confident that their findingsrepresent a nationwide perspective.

Page42


43/57

EXECUTIVE 2: Is there any other potential for this product?

JIM: Actually, theres a feelingeven from some of the business ownersthat consumerswould be interested in a similar product.

EXECUTIVE 3: I hear there is a product like that, a new consumer credit card that offersrewards, on track for approval now.

JIM: Well, then, we have an opportunity to coordinate our efforts. Let's defer that point for now.

EXECUTIVE 1: Jim, I believe that this product could really grow our small business segmentand you have my support. But youd better move quickly, because Mary will be making her

presentation to the Board within a few months, and were all aware that only one product will getthe years big marketing push.

EXECUTIVE 2: You have my support as well.

EXECUTIVE 3: And mine.

FEBRUARY 4

PHONE CALL BETWEEN JIM WILSON AND MARY MONROE

JIM: Hi, Mary! Do you have a few minutes?

MARY: Just a few, Jim. Whats up?

JIM: I'd like to submit a product for your consideration based on an analysis that my team hasdone. It's a card for small business owners that will allow themonce they qualifyto purchasecapital items. The card has a credit line of $200,000, and it also has a reward component.

Page43


44/57

MARY: I'll be happy to review the analysis, but I will warn you that Ive had an independentmarketing firm perform a thorough statistical analysis and their findings are for a consumer

product that offers rewards.

JIM: I'll send you the analysis. It uses focus groups with clients and small business-owners, aswell as statistics. Its fairly robust.

MARY: Like I said, I'll be happy to review it, but Ill have trouble disputing the conclusion fromthe Best Market. Ive reviewed the report with several executives and they support the consumer

product as well. Were planning to recommend it.

JIM: Mary, I also got feedback from many executives. They encouraged me to contact youimmediately, because they have a lot of faith in this product. I know time is short, but consider that this card could span into the consumer market as well.

MARY: Send me the analysis.

FEBRUARY 4

E-MAILFROM: Mary MonroeTO: Jim Wilson

RE: New product analysis

Thanks for getting the data to me so promptly, Jim. Ill look it over and let you know what Ithink.

FEBRUARY 10OFFICE MEETING BETWEEN BRIAN ALLEN AND MARY MONROE

MARY: I held the meeting with the executives, as you suggested, and they agree with Best

Page44


45/57

Market that we should move forward with this product. Let me go through the material with youas I did with them in the meeting.

BRIAN: That's not necessary. I've looked at the results.

MARY: Can I help with this any further?

BRIAN: Mary, you know our new products have pretty much tanked in the past couple of years.Were under a lot of pressure to deliver something that will be profitable and increase marketshare. Is this going to do it?

MARY: I have no reason to doubt Best Markets results. Theyre a prestigious firm with aninternational presence. Id like to begin putting together a formal presentation for you. Should weuse the same format as last year?

BRIAN: Start thinking about it, but dont go all out just yet. I still havent made up my mindabout this product.

MARY: We don't have a lot of time. Do you have specific concerns that I can help with?

BRIAN: No, I just need some time to consider this. Ill let you know when I decide.

MARY: Okay. In the meantime, I'll only look for last year's submission and start creating aformat for this year.

BRIAN: Thanks for your review, Mary. Ill be in touch soon.

FEBRUARY 11

E-MAIL

FROM: Jim WilsonTO: Brian AllenRE: Small Business Card

Brian,

Page45


46/57

Id like to request a brief meeting to discuss a new product with great potential: a small-businessowner card with a reward component. Ive met with my executive staff and they are totallyonboard with this. In fact, they encouraged a recommendation to you and Mary. Ive alreadyspoken to Mary, and she is reviewing my teams analysis, but Id like to share it with you aswell.

I recognize that you are very busy and am willing to meet with you as soon as your schedule permits. I think this is that important.

Regards,Jim Wilson

FEBRUARY 11

E-MAILFROM: Brian AllenTO: Jim WilsonRE: Small Business Card

Please come to my office next Tuesday at 10 a.m. to discuss this potential new product.

FEBRUARY 15EXCERPT FROM MEETING BETWEEN BRIAN ALLEN AND JIM WILSON

JIM: So thats it in a nutshell. The results of the focus groups are conclusive; there is a definiteneed for a new product that allows small-business owners who qualify to have a high-limit card

that can be used to purchase capital items. What do you think?

BRIAN: Im concerned that Mary isnt completely onboard with this. Shes the one whosultimately responsible, with my approval, for the submission of the new product.

Page46

dteal

cusups

l beected


47/57

JIM: Brian, I wouldn't have come to you without telling Mary. But I'm here because theexecutives I shared this with strongly encouraged me to push this product. They see great

potential beyond even the small business owners. Im very concerned, based on the data mymanagers collected, that if we don't do this, were not just sacrificing new profits, but old ones;my managers think that their customers will leave us for banks that are offering what they need.Even the executives at my meeting have been hearing rumblings from small business ownerslooking for new banking opportunities. Weve been the leader in our industry; can we afford tolose this position and possible market share?

BRIAN: I hear what youre saying, Jim, but I need to review the results carefully. Give me sometime to think about it; I wont take too long.

MARCH 1

E-MAILFROM: Mary MonroeTO: Jim WilsonRE: New product decision

Jim,

Brian asked me to let you know that a decision has been made to recommend the consumer product to the Board. I hope you will support this decision.

Regards,

Mary

MARCH 9EXCERPT FROM MEETING BETWEEN BRIAN ALLEN AND MARY MONROE

MARY: Heres the Best Market report Ive prepared for the Board. Lets go over it to make sureyou dont have any issues or concerns.

Page47


48/57

BRIAN: If nothings changed, then Im okay with it. We dont have to go over it all again.

MARY: Fine. Ive also included a copy of the survey and results if the Board wants to see that.Should I go over them?

BRIAN: Let's go over a few.

MARY: Good. Here is a question we asked: "How likely is it that you will sign up for InstantRewards if it only includes your top-ranked benefit?" The results show that they would chooseInstant Rewards. Here's another: "How likely is it that you will change banks to join Instant

Rewards if it only includes your top three ranked benefits?" The results indicate that we woulddraw new customers. Im telling you, this is a winner.

BRIAN: I dont disagree, but I am still concerned that launching this product will be a slap inthe face to our small business owners, and that were risking losing them. This is going to be aconcern for the Board as well.

MARY: We don't see the small business card as a priority.

BRIAN: I hope you're right. Thank you for putting this together. Ill review the rest and let youknow if I have any questions.

MARCH 10E-MAIL

FROM: Alexis Andrews

TO: Board of DirectorsRE: New Board member

Please join me in welcoming Bea Hansen to the USA World Bank Board of Directors. Beacomes to us from the world of academia, where she is professor of statistics. She also has had avery bright career as an author and consultant in the area of statistics. We are very fortunate to

Page48


49/57


50/57

BRIAN: Sure, I have the data right here; give me a second to find it. Okay, here it is. The samplesize was about 140,000, and I will have to get back with you on the time frame.

BEA: That's a large sample. How was the data collected?

BRIAN: When consumers log on to our Web site, theyre asked to fill out a survey. Wecollected more than 140,000 observations that way.

BEA: I'm not so sure the sample is representative; the gender percentages look wrong. You cancheck on the Census Bureau Web site, but I believe there are slightly more females than males inthe population. Youre showing a higher percentage of men in this sample, possibly because menare more comfortable with computer technology and therefore more likely to respond to an

online survey.

BRIAN: I can appreciate what you're saying, but isn't it true that with such a large sample anyerrors in the data will randomize out?

BEA: Random error, yes, but not sampling error. Have you ever heard about the classic screw-up in the Literary Digest presidential poll of 1936? Literary Digest was a very reputable, high-

brow magazine that confidently predicted that Alf Landon would be elected in a landslide. Theyhad a sample size of more than two million people. George Gallup predicted the opposite; hesaid that FDR would win in a landslide. His sample was only about 1,500. So how could

Literary Digest, with a sample of two million be so wrong? Because they collected their samplefrom phone books and auto registrations. Who owned a car or had a phone in 1936? My parentsdidn't. My father was 22 years old, working on the WPA, with one small child at home and asecond on the way. 1936 was the only time in his life that he voted Democratic. In 1936, andtoday, for that matter, Democrats tend to be more downscale.

BRIAN: So youre doubting these results?

BEA: I would be a lot more confident with a truly random sample. I mention age and income because its well-known that tech-savvy people tend to be younger and more upscale.

I also would like that timeframe data. If I may, I'd like to bore you again with another historicalanecdote, this one about the 1948 election. Remember the classic photograph of Harry Trumanholding up the Chicago Tribune with the headline, "Dewey Defeats Truman"? New York

Page50

mplin

thods

plyicbabil

cept

ilitat


51/57


52/57


53/57

JIM: Sure can, sir. Ill have it to you in a couple of hours.

BRIAN: Thank you, Jim.

JIM: By the way, I agree with that we could use a consistent methodology for launching new products. Once were through this years launch, Id be interested in chairing a committee to thatend. See you in two days!

APRIL 12EXCERPT FROM EXECUTIVE BOARD MEETING, ALEXIS ANDREWS PRESIDING

BRIAN: If I may, I'd like to pick up where we left off last time.

BEA: Before you do, Brian, I want to point out that your huge sample size may be distorting theresults. A significant correlation, such as you found, means that the coefficient, or weight of themathematical model, is unlikely to equal zero. It means that there is a relationship between ageand the likelihood of purchase and the likelihood of changing banks. The information about agehelps us estimate purchase and bank changing.

How much does it help? This is where the R-squared measure you included comes in. R-squaredis a ratio of how much of the variation in the likelihood and change scores is accounted for byage. Another way to look at this is to think about how much our ability to predict a bank changeimproves if we know age. Not very much for allonly three or four percent. This small amountis "statistically significant," but because of the huge sample size, it has little practical value. Wehave to use some common sense here.

Even predicting the bank change with the top three benefits is marginal. I'm not sure I wouldrecommend spending millions of dollars marketing this product to young people when ageincreases our ability to predict by less than 20 percent. We need to look at other variables. Weneed to get the R-squared up by quite a bit.

There is an old saying in my business that an R-squared between 70 and 80 percent is a goodworking model. R-squared between 80 and 90 percent is excellent. R-squared between 90 and 95

percent qualifies the researcher for the statistician's hall of fame. And an R-squared above 95 percent means that you are probably lying.

Page53

ply

rmaltribon

ilita

sine

cisio

kinghapte,7)

plynfidce

-7


54/57

BRIAN: If we could just...

BOARD MEMBER 1: Something doesn't seem right here. I didn't know that you could do percentages on ranked data. Im confused, Dr. Hansen.

BEA: You're right about trying to divide ranked data. It is meaningless. But Brian is giving us percentages of the average ranks, which is fine because this is ratio data.

Let me back off a bit. There are four levels of measurement: nominal, ordinal, interval and ratio. Nominal is just a name, like the numbers for football players. It is meaningless to say that

number 42 is twice as good as number 21.

Ordinal data is like the finishers in a marathon. The first and second place people could beseconds apart, while the third guy could be miles back. With ordinal data, there is an order to thenumbers but they are not equally spaced. There is no equal interval, so division is meaningless.

Interval data like the Fahrenheit temperature scale has equal intervals, but the zero is just another point on the scale and doesn't signify the lack of any heat. The difference between 20 degreesand 40 degrees is the same amount of heat as the difference between 40 degrees and 60 degrees.That's equal interval. But 40 degrees is not half as hot as 80 degrees. Division is meaningless

because zero degrees Fahrenheit is not absolute zero.

And finally, ratio data, like money, has both equal interval and a meaningful zero point.Addition, subtraction, multiplication, and division are all possible.

You brought up a very good point. It is important to remember that when talking about anindividual respondent's rankings, we cannot say that their top-ranked benefit is twice asimportant as their second-ranked benefit, and three times as important as their third-ranked

benefit. There is no equal interval between the ranks. Like the finishing places in a race, one and

two could be millimeters apart, while third is miles back.

BOARD MEMBER 2: I just noticed something in the questionnaire. Questions IR 3, 4, and 5ask about both signing up for the Instant Reward card and about changing banks. It seems likethese two issues should be in separate questions, doesn't it?

Page54


55/57

ALEXIS: It seems obvious that we need to clarify a couple of issues before the Board can makeany decisions about launching the Instant Rewards product. Bea, will you set up a meeting withBrian and Mary to work out the kinks in this research report?

BEA: Tuesday or Wednesday mornings next week look pretty good right now. Brian, can I giveyou a call on Friday to set something up?

BRIAN: That sounds good. I can juggle some things and make ourselves available.

ALEXIS: For the time being, then, let's consider this meeting adjourned.

APRIL 13MEETING WITH BRIAN ALLEN, MARY MONROE, JIM WILSON, AND TOMARAYA

BRIAN : Well, I don't ever want to go through that again. If we have to redo the Instant Rewardsresearch, it will take months. We need to have something for the Board, just in case. Jim, explainagain how the Small Business Card research was conducted.

JIM: We used the same focus group method that we've always used in the past. Is it perfect? Idon't know. I shared some of the details with Tom earlier.

TOM: I wonder if Dr. Hansen will bring up what is known as "Pygmalion in the Classroom." Itwas a study in the 50s where teachers were told that a few of their incoming students would"bloom" during the next school year. The children that were supposed to bloom were justrandomly selected, but by the end of the school year, these kids actually had made great stridesin their school work. How the teachers treated these special students seemed to be the key.This wasn't a fluke; the research has been replicated hundreds of times on everything from lab

rats to college sophomores. Dr. Hansen might point out that personal bankers could have been,inadvertently, leading the customers in the direction the personal bankers wanted them to go.

BRIAN: How can we fix this?

MARY: We could have Best Market Research conduct focus groups of our business customers.

Page55


56/57

BRIAN: How long will it take?

MARY: If we're willing to pay for a rush job, three or four weeks, maybe.

BRIAN: At this point, I don't really care about the money.

Sampling methods (chapter 8)

TOM: We need to make sure that we get a sample that represents our customers. Even if we justcount the number of positive and negative comments about the product, Dr. Hansen will

probably play the random sample card again. Jim, what do you think are the crucial things insmall business decisions?

JIM: Over the years, the personal bankers have told me that the first priority is the size of thecompany. Second would probably be whether or not they were the founder of the company.Founders tend to be more emotionally involved in their companies than professional managers.

BRIAN: So we need to make sure that the companies the market research company interviewsrepresent a full range of sizes and a good mix of founders and CEOs who came into the businessafter it had started.

TOM: Yes, we can have Best Market Research use these two variables to stratify the sample torepresent our customer base and then randomly select companies within each of these cells to bein the focus groups.

JIM: As I recall, we interviewed 75 or 80 customers. Will that be a large enough sample tosatisfy Bea?

MARY: We'll check with Best Market Research, but that seems like a good amount.

TOM: But even in focus groups, we have to watch the response rate. If only 25 percent of thecompanies are willing to take part in the focus groups, Dr. Hansen might say that they somehoware different from the ones who refused. I don't know what the response rate should be in focusgroups, but in survey research, the rule of thumb is that a 50 response rate is consideredadequate, 60 percent is considered good, and 70 percent is excellent.

Page56

mplin

thod


57/57

BRIAN: While I've got you here, I want you to start thinking about something Alexis mentionedthe other day. She was wondering where we got our ideas for new products. I told her that the

bank presidents and the personal bankers have their ears to the ground. They hear things.

JIM: Harry Truman used to say that some politicians spend so much time with their ears to theground that all they hear from are the grasshoppers.

BRIAN: So do we need to be more systematic about how we gather new ideas? Mary and Jim,would you start talking to your colleagues about how other companies in and out of bankingthink up new products? Another thing is that there seems to be a lot of bickering about consumer versus commercial products. Our Board of Directors only wants to launch one product a year.

JIM: Have they ever said why only one?

BRIAN: If there ever was a reason, it has been lost to the ages. I want Jim and Mary to startlooking into something. We need to be able to launch both commercial and consumer products atthe same time. Maybe we need to separate our market research function into two branches so thatthe two of you can explore opportunities separately.

MARY: Won't the Board say that we can't afford to do more than one product a year?

BRIAN: Probably. We need to show them that both can be profitable. I'm going to talk to theaccounting and IT people about setting up a system to estimate payback period or ROI or somegood measure like that.

JIM: You probably also want to include the finance department.

BRIAN: Good point. Mary, Bob and I have to deal with Bea Hanson next week. Mary, you andJim get together to set up the Small Business Card focus group project. I'll talk to finance,

accounting, and IT about ROI measures. Mary and Jim, start poking around about systems for gathering new product ideas. It is going to be a tough couple of weeks. Jim, Mary, and I willmeet again on Friday at 8 a.m. to see where we stand and what barriers need to be overcome.

dte

parat

Documents

Managerial decision making (USA World Bank)