Market Research

Marketing Research Project Wendy’s: History and Life after Dave Thomas

Submitted By: Group-7 Members: Rahul Budhia, Rajnish Sharan Singh, Ranjana Mohan, Ravi Gupta, Richa Choudhary, Rishabh Kant Singh Group-8 Members: R. Sandhya, Sabyasachi Bhanja, Sandarbh Goswami, Saptorshi Bagchi, Sarang Pious, Sarath Chandra

Chapter 1 Q1. Discuss the role marketing research can play in helping a fast-food restaurant such as Wendy’s formulate sound marketing strategies. In a highly competitive market similar to that of a restaurant business, each player has to keep a track on the preference movements of the consumer and also have to closely watch what the competition is doing. For each player there is only one chance to proactively implement a strong positioning before anyone else does it. Therefore there arises a high demand of understanding the market before others. Wendy’s is a distinctly positioned brand in the perception map of the user base, but due to changing lifestyles, eating habits and choices change. McDonald’s and Burger King with their huge capital and global presence have the capability of tapping this problem before Wendy’s can do. Therefore, there is a tenacious need for Wendy’s to conduct market research to safeguard the position they have in the mindshare of the consumers. Through research, the management can identify the key points to service the customers with best offerings. This will help them design an apt marketing strategy. Awareness of competition and our brand: Through a smart research design we can find what our consumer feels and knows about all the competing brands in the segment; how they react to the advertisements/schemes/offers coming from the market; how adaptive they are to the changes coming in the products/services offered by the brands. Questions like this can be easily answered through research. The results will enable the management to use the information as a tool against difficulties coming from the market. Know what customer needs: Marketing doesn’t end with a good branding and enhanced visibility. It also includes finding out what is missing and what should be added in the offerings. With this research we can effectively find out what is lacking in the menu which the target consumer looks for and bring it on the menu before any one in competition does. Customer demographics: Having consumer base’s explicit details will certainly ease the process of designing the marketing strategy. Managers would know exactly where to invest and how much to invest to grab more eyeballs and thus big digits in the balance sheet.

Dining Preferences: With every 100 miles (it is said that) people differ in personal behavior and preferences. To prove it, the sales of different stores in different geographical locations with same offerings and identical infrastructure are different. With our research we can find out from the masses what are the developing tastes of dining in that particular geographical area. This will help us redesign the service encounter for the users and change the platform of perhaps ‘eat-in’ to ‘drive-in’ as an instance. Receptiveness v/s Culture: Local culture also plays a major role in the way how people react to communications floated by the brands. States like Alabama are more family-oriented and relaxed in nature, whereas people in Colorado are fast moving and career oriented. If we happen to understand them more in detail, we can easily design our advertising more effectively.

Chapter 2 Q1. Wendy’s is considering further expansion in United States. Define the management decision problem. Management Decision Problem: “Which geographic locations in the United States should Wendy’s expand?” Reason- For further expansion it is necessary to know which location Wendy’s should cater to as it would require analysis on aspects such as cost of setting up a store , number of competitors, total population in that geographic region. Q2. Define an appropriate marketing research problem based on the management decision problem you have identified. Marketing Research Problem: Determine the customer preference towards consumption of fast food in the determined location with respect to the competitors and the manner in which they would like to be served. Reason- Based on the management’s decision to open a store it then becomes important to understand the market situation in that region so as to create an efficient marketing plan for its customers for appropriate positioning, featuring on the price points of their offering, their target population and the mode of service delivery preferred.

Chapter 3 Q1. Formulate an appropriate research design for investigating the marketing research problem you have defined. RESEARCH DESIGN FOR WENDY’S EXPANSION PLAN: Research design is a blue print of the research we wish to conduct as a company, to find out the reasons of a failure or possibilities of progress. To understand whether Wendy’s will prosper or not by expanding their stores PAN USA, we need to have a sharp research design. Following illustration defines the plan:

Chapter 4 Q1. Use the Internet to determine the market shares of the major national fast-food chains for the last calendar year. Market Share of leading fast food chains in the United States

Mkt.Share in % of Major Fast Food Chains in U.S in 2013 Mc Donald's Corporation

21.7

Yum Brands Inc

8.1

Doctor's Associates Inc.

6.7

Wendy's International Inc.

5.5

Burger King Corporation

4.1

Other

53.9 0

10

20

30

40

50

60

Mkt.Share in %

Major Fast Food Chains in U.S Mc Donald’s Corporation. Yum Brands Inc. Doctor’s Associates Inc. Wendy’s International Inc. Burger King Corporation Other

% of Mkt. Share in 2013 21.7 8.1 6.7 5.5 4.1 53.9

McDonald’s held, by far, the largest market share of the United States fast food industry in 2013. Its closest competitor was Yum! Brands - owner of popular chains Taco Bell, KFC, Pizza Hut and Wing Street. The top five brands accounted for just under half of the entire U.S. fast food industry, which, in 2013, generated over 191 billion U.S. dollars in revenue. This revenue was forecasted to rise above 210 billion dollars in 2018.

Q2. What type of syndicate data will be useful to Wendy’s ? Type of syndicate data to be used by Wendy’s could be taken from consumers who is an important stake holder in a fast food industry. The methodology to be followed is to conduct surveys encompassing 1) The life style and psychographics of the consumers.-This would cater to the demographics and preferences of the target population and further present a clear view in the segmentation process. 2) The advertising evaluation of Wendy’s competitors.- Since Wendy’s competitors always had an edge over advertising Wendy’s needs to work on an appropriate advertising plan for efficient positioning among the target consumers.

Chapter 5 Q1. Discuss the role of qualitative research in helping Wendy’s expand further in the United States. Wendy’s is one of the leading fast food companies in United States with over 10000 outlets. Although the company has substantial market share after McDonald’s and Burger King, it has to improve its capacity to compete with these rivals. In order to further expand in United States Qualitative research will be an important tool of survey. Wendy’s can select a group of 6 to 10 customers of different age groups for a Focus Group Discussion. The moderator should conduct the discussion with an aim to achieve the following: 1) To get a clear image of fast food industry in United States 2) To understand the customers’ perception towards other competitors (McDonald’s and Burger King). This will help the company identify the area of expansion 3) To understand the customers perception towards its product and services. The qualitative study will give a theoretical background and enhance the idea about customers’ perception of the brand

Chapter 6 Q1. Wendy’s has developed a new fish sandwich with a distinctive Cajun taste. It would like to determine consumer’s response to this new sandwich before introducing it in the marketplace. If a survey is to be conducted to determine consumer preferences, which survey should be used and why?

In order to determine the consumer preference for Wendy’s newly developed fish sandwich mall intercept form of personal interview survey should be used. As in this form of survey the respondents are intercepted while they shop in the shopping mall and they are tested. As Wendy’s is introducing its new fish sandwich it will like to know that what is the consumer review on this sandwich and how consumer take this sandwich and so this type of survey method will be best to take out their taste view their size, sample, price view for this sandwich. Another important benefit of using this type of survey is that it will not only give consumer’s preference but also it will act as a test market on its own and that too here consumers are not paying for anything. As the survey is about food so physically testing the food is very important and this survey method will give all that benefits and tasting of food will be proceeded by small questionnaire in which customer can show their level of satisfaction or dissatisfaction for this newly developed fish sandwich. Given below are the fact and findings from various researches to substantiate the importance and advantages of using mall-intercept survey methodA. Background Mall intercept surveys are widely used and (theoretically) able to reach a large segment of the population. In any given two-week period, about 2/3 of U.S. households shop one or more times at a mall. According to a CASRO membership survey, about 25% of all marketing research and 64% of personal interviews are conducted at malls. B. Pluses and Minuses The good things about mall samples are: 1) Experimental control. 2) Ability to see things. 3) Availability of kitchens, etc. 4) Minimal Cost. C. Effect of Mall Samples on Results 1) For copy, concept, and product tests, data suggest that mall samples understate scores. 2) Ossip reported four studies that found lower top box concept scores for mall surveys compared to door to door, even after controlling demographic differences. 3) Gannon reported study comparing mall and mail panel for a concept/product test. Mall study got lower concept top box but higher product test attribute ratings. D. "Ideal" Mall Sampling Plan

According to an article by Seymour Sudman, to achieve a very good sample via the mall intercepts method. However, this is what can be done. 1) Randomly select states or regions. 2) Randomly select cities within region. 3) Randomly select malls within cities. 4) Post interviewers at randomly selected mall entrances. 5) Interview all days and all times mall open. 6) Count traffic so interviews are proportional to traffic based on day of week and time of day. 7) Determine frequency of mall shopping and weight sample so that frequent shoppers not overrepresented.

Chapter 7 Q1. Discuss the role of experimentation in helping Wendy’s determine its optimal level of advertising expenditures. Wendy’s is the third hamburger chain by sales after McDonalds and Burger King. Although having a major market share, it needs to brand itself in such a way that people are more drawn towards its quality and optimal price. Wendy’s has introduced various new meals in their list which emphasizes on higher quality, great taste and fresh and never frozen ground beef. So in order to study the customer’s awareness of the competitors and how they respond to the new meals, Wendy’s can perform standard test marketing for their new meals by introducing them to customers and collecting reports of what they thought about the new meals. And if the initial findings are found successful, they can expand the same test to different cities and also ask whether they would like to have any changes in the taste or quality or price and act accordingly. It can also help in determining how the consumers rank Wendy’s in comparison with its competitors.

Chapter 8 Q1. Illustrate the use of primary type of scales in measuring consumer preferences for fast-food restaurants.

Scale

Basic characteristics

Nominal

Numbers identify and classify objects Numbers indicate the relative position of objects but does not indicate the relative difference in their magnitude Difference between objects can be compared Zero points is fixed, ratios of scale can be computed

Ordinal

Interval Ratio

Consumer preference measurement use Dish name, numbering of dish, Food categories, Store name Preference ranking, quality ranking, customer satisfaction ranking, food quality ranking, service quality ranking. Attitudes, opinion, index numbers, age group, Income group Age, income, cost, frequency

Chapter 9 Q1. Illustrate the use of likert, semantic differential, and staple scales in measuring consumer preferences for fast food restaurants. Answer: LIKERT SCALE : This scale is a widely used rating scale that requires the respondents to indicate the degree of agreement or disagreement with each of a series of statements about the stimulus object. Disagree completely

Disagree somewhat

Neither Agree Agree nor Disagree somewhat

Choose one response for each statement 1) I try to stay current on the latest health and nutrition information.

Agree completely

2) I read nutritional labels on most products I buy. 3) I consider the amount of fat in the foods I eat at fast-food restaurants.

Semantic differential: This is the 7-point rating scale with endpoints associated with bipolar labels that have semantic meaning. Please rate the restaurants you, yourself, have eaten from in the past three months using a 7-point scale, where 7 means you think it is perfect, and “1” means you think it is terrible. Terrible (1)

2

3

4

5

6

Perfect (7)

Stapel Scale: A scale for measuring attitudes that consists of a single adjective in the middle of an evennumbered range of values, from -5 to +5, without a neutral point (zero) FAST-FOOD RESTAURANT +5 +4 +3 +2 +1 High Quality -1 -2 -3 -4 -5

Chapter 10 Q1. Develop a questionnaire for assessing consumer preferences for fast-food restaurants.

Questionnaire for assessing consumer preferences for fast-food restaurants. 1. Age

2.

3.

4.

5.

6.

7.

8.

 5-12  13-16  17-23  24-32  >32 Gender  Male  Female  Other Ethnicity  Asian  European  American  South American  African  Australian Occupation  Student  Working Professional  Other What is your annual income?  Below 2 lakh  2 lakh – 5 lakh  Above 5 lakh Relationship status  Single  Recently married  Married and staying with children  Married and staying with children How often do you visit a fast food restaurant?  Daily  Once a week  Occasionally  I don’t like fast food Rate the following in a fast food restaurant which would influence you to visit a fast food restaurant from 1 -5 (1 being the lowest and 5 being the highest)

 Food  Ambience  Comfort  Cost effectiveness  Healthy food  Quick service 9. Which food are you likely to order in a fast food restaurant?  Chicken Burgers  Beef Burgers  Ham Burgers  Potatoes  French fries  Sandwiches  Shakes  Deserts 10. With whom do you usually like to visit a fast food restaurant?  Family  Friends  Someone special  Alone

Chapter 11 & 12 Q1.What sampling plan should be adopted for the survey of chapter 6? How should the sample size be determined? Group7Ans. Wendy’s has developed a new fish sandwich with a distinctive Cajun taste. It would like to determine consumers’ response to this new sandwich before introducing it in the marketplace. We can take a convenience sample of 200 visitors (approx equal male- female ratio) asking them about what level of satisfaction they get from this new offering; are they satisfied with the quality, price and quantity level of this new offering?

Group8Ans. Since we are considering mall intercepts as one of our sampling techniques, it would fall under the category of probability sampling. Since we have no control over which customer to be considered, the probability of every customer to be chosen would be equal.

Typically for testing new products, a minimum sample size of 200 is considered while the range of sample size would be between 300 and 500. The appropriate sample size of a particular study can be determined using both qualitative and statistical factors. The qualitative factors include the importance of the decision, the nature of the research, the number of variables, the nature of analysis, resource constraints etc. The statistical approach would involve determining the sample size based on the construction of the confidence intervals around sample means and proportions. "Ideal" Mall Sampling Plan1) Randomly select states or regions. 2) Randomly select cities within region. 3) Randomly select malls within cities. 4) Post interviewers at randomly selected mall entrances. 5) Interview all days and all times mall open. 6) Count traffic so interviews are proportional to traffic based on day of week and time of day. 7) Determine frequency of mall shopping and weight sample so that frequent shoppers not overrepresented.

Chapter 13 Q1. How should the fieldworkers be selected and trained to conduct the fieldwork for the survey? Ans. A survey is a method of descriptive research design which is in turn, a conclusive research methodology. The purpose of such a technique would be to arrive at a conclusion so as to address a problem. In order to address the marketing research problem that has been defined, we probe into what a survey actually means. A survey is a structured questionnaire given to a sample of a population and designed to elicit specific information from respondents. The field force is made up of both actual interviewers and supervisors involved in data collection. Since a survey involves less interaction except for interviews, requirement of such personnel is limited. However, there exists a potential for bias in (1) selecting respondents – selecting the incorrect sample (2) asking questions – omitting certain questions (3) recording answers – recording incorrectly or incompletely. Interviewers can influence the bias in their own ways – inflection, tone of voice, suggesting answers, etc.

Hence, while selecting the fieldworkers, care should be exercised to avoid the above mentioned possibilities which might flaw the research of hamper its results. In a computer based or internet survey, such occurrences are low. Hence a team of supervisors must be selected to train them and to supervise the interview process. If interviews are conducted across geographies, the scope of such supervision is limited.

Chapter 14 Q1. How should the missing values be treated for the following demographic variables : education (D5), income (D6), employment status (D7), and marital status (D8) ? Variable D5 Variable D5 is a categorical variable which shows the different levels of education. The unanswered responses would be where the respondent answered “Prefer not to answer”. This can be considered as a missing value and it can be replaced with the most frequently occurring ‘level of education’ response to this question. Variable D6 Variable D6 is a categorical variable which shows the respondent’s family’s annual household income level. The unanswered responses would be where the respondent answered “Prefer not to answer”. This can be considered as a missing value and it can be substituted by an imputed response. The respondent’s pattern of responses to other questions is used to impute or calculate a suitable response to the missing values. Variable D7 Variable D7 is a categorical variable which shows the ‘employment status’ of the respondent. The unanswered responses would be where the respondent answered “Prefer not to answer”. This can be considered as a missing value and it can be replaced with the most frequently occurring ‘employment status’ response to this question. Variable D8 Variable D8 is a categorical variable which shows the ‘marital status’ of the respondent. The unanswered responses would be where the respondent answered “Prefer not to answer”. This can be considered as a missing value and it can be replaced with the most frequently occurring ‘marital status’ response to this question.

Q2. Recode payment method (D1) by combining Debit card, Check and other into one category. Variable D1 has been recoded by combining Debit card, Check, and other into one category Old values Cash- 1 Credit card- 2 Debit card- 3 Check- 4 Other- 5 Prefer not to Respond- 6 New values Debit, check, other- 1 Cash- 2 Credit card- 3 Prefer not to Respond- 4 Q3. Recode number of people living at home (D3A) as follows: For adults age 18+, four or more should be combined into one category labeled 4 plus; for each of the three remaining age groups ( under5, 6-11, and 12-17), two or more should be combined into a single category labeled 2 plus. Variable d3a_1 has been recoded into d3a_1_r. New values in d3a_1_r 0 is 0 1 is 1 2 is 2 3 is 3 4-15 is 4 plus Variable d3a_2 has been recoded into d3a_2_r New vales in d3a_2_r 0 is 0 1 is 1 2-9 is 2 plus Variable d3a_3 has been recoded into d3a_3_r New vales in d3a_2_r 0 is 0 1 is 1

2-9 is 2 plus Variable d3a_4 has been recoded into d3a_4_r New vales in d3a_2_r 0 is 0 1 is 1 2-9 is 2 plus Q4. Recode education (D5) by combining the lowest two category and labeling it completed high school or less. Variable d5 has been recoded into variable d5_r New values for d5_r 1 - Completed high school or less 2- Some college 3- Completed college 4- Post graduate 5- Prefer not to answer Q5. Recode income (D6) by combining the highest three categories and labeling it $100,000 or more. Variable d6 has been recoded into d6_r New values for d6_r 1- Under $25000 2- $25000 but under $50000 3- $50000 but under $75000 4- $75000 but under $100000 5- $100000 or more 6- Prefer not to answer Q6. Recode employment status (D7) by combining homemaker, retired and unemployed into a single category. Variable d7 has been recoded into d7_r New values for d6_r 1- Home maker, retired, unemployed 2- Full time 3- Part time 4- Student 5- Prefer not to answer

Q7. Classify respondents into light, medium and heavy users of fast food based on the frequency distribution of S3A: In the past four weeks, approximately how many times, have you, your-self, eaten food from a fast-food restaurant? Use the following classification: 1-4 times =light, 5-8 times= medium, 9 or more times= heavy. Variable s3a has been converted into s3a_r New values for s3a_r 1- 1 to 4/ light 2- 5 to 8/ medium 3- 9 to 99/ heavy

Chapter 15 Q1. Run a frequency distribution for all variables except respondent ID (responseid). Why is this analysis useful? Ans. This analysis is useful because we obtain a count of the number of responses associated with different values of one variable and to express these counts in percentage terms. For the frequency distribution please refer the output file. Q2. Cross-tabulate fast food consumption classification (recoded S3A) with the demographic characteristics: age(S1), gender(S2), payment method(D1), number of people living at home(D3A), education (D5), income (D6), employment (D7), marital status (D8), and region. Interpret the results Ans. It is clearly seen from the crosstabs that young people in the age category of 18-24 have the maximum consumption and as the age increases, the frequency of consumption decreases. There are total 575 respondents in the age category of 18-24. Out of these, 90 respondents (max) visited the fast food restaurants 4 times in past one month. Out of 1440 respondents, 1277 respondents had done the payment by cash. College goers visits fast-food restaurant more as compared to graduates and post-graduates.

Q3. Cross-tabulate payment method (recoded D1) with the demographic characteristics: age(S1), gender(S2), number of people living at home(D3A), education (D5), income (D6), employment (D7), marital status (D8), and region. Interpret the results. Ans. According to age category, around 88.6% respondents had done payment with cash and rest done with the credit card(3.2%) and debit card(6.5%). There are 85% respondents in each category of annual household income which had done their payments with cash which was followed by debit card. Q4. Cross-tabulate eating there more often, less often, or about the same as a year or so ago(q8_1,q8_7,q8_26,q8_36,q8_39) with the demographic characteristics : age(S1), gender(S2), payment method(D1), number of people living at home(D3A), education (D5), income (D6), employment (D7), marital status (D8), and region. Interpret the results. Ans. Out of four brands (i.e. Mc Donald’s, subway, Burger King, Wendy’s), Mc Donald’s have the highest number of responses. Age category

More often

About the same

Less often

18-24

Subway>Mc

Mc Donald’s>burger Mc

Donald’s>Burger

King>Subway>Arby’s King>Subway>Arby’s

Donald’s>burger

King>Arby’s 25-29

Subway>Mc


Donald’s>burger

Donald’s>Burger


King>Arby’s 30-34

Subway>Mc


Donald’s>burger

Donald’s>Burger


King>Arby’s 35-39

Subway>Mc


Donald’s>burger

Donald’s>Burger


King>Arby’s 40-45

Subway>Mc


Donald’s>burger

Donald’s>Burger


King>Arby’s In all the age category’s, number of responses for Arby’s is least, so people preferred going to the other brands than Arby’s. Number of female respondents is greater than male respondents in all categories i.e. More often, About the same and less often. Number of responders paying with cash is high as compared to other mode of payment in all categories i.e. More often, About the often and less often. The customer who had completed their high school found to be visiting more often to the Arby’s as compared to other categories. For the category of some college and completed college, they contributed highest proportion to the total customer but they are losing out their customer base as compared to previous year. Customers earning between 25,000-50,000 and 50,000-75,000 contributes highest. They are losing out customer base of earning income of 50,000-75,000 who visits the least compared to last year. South region contributes highest proportion of customers. They are losing out on customer base of northeast when compared to last year. Q5. Do the ratings on the psychographic statements (q14_1, q14_2, q14_3, q14_4, q14_5, q14_6, q14_7) differ for males and females (S2)? How would your analysis differ if the ratings on the psychographic statements were treated as ordinal rather than interval scaled. Ans. Yes, it differs Gender wise. Female are more concerned about latest health and nutrition information, they prefer buying product with nutrition levels, make more effort to find out the nutritional content of the food they eat and consider the amount of fat in their as well as in their kids food. The analysis won’t differ when we take data in interval scaled. Q6. Do the respondents agree more with “I have been making an effort to look for fast-food choices that have better nutritional value than the foods I have chosen in the past” (q14_6) than they do

with “I consider the amount of fat in the foods my kids eat at fast food restaurants” (q14_5)? How would your analysis differ if these ratings were treated as ordinal rather than interval scaled? Answer:

Above frequency table shows two variables: Variable 1: “I have been making an effort to look for fast food choices that have better nutritional value than the foods I have chosen in the past” Variable 2: “I consider the amount of fat in the foods my kids eat at fast food restaurants”

Both these variables are measured on interval scale and the percentage of respondents that completely agree with 1st statement is higher than the 2nd statement. If these variables are changed on ordinal scale, the analysis would not differ.

Chapter 16: Q1. Do the restaurant rating (q9_1,q9_7,q9_26,q9_36,q9_39) differ for the various demographic characteristics (some recoded as specified in Chapter 14) : age (S1), gender (S2), payment method (D1), number of people living at home (D3A), education (D5), income (D6), employment (D7), marital status (D8), and Region. Interpret the results. 1) Variable q9_1 An N-way anova has been done by taking variable q9_1 (Arby’s rating) as the dependent variable and s1, s2, d1, d3a, d5, d6, d7, d8, region as dependent variables. The following results were observed: Levene's Test of Equality of Error Variancesa

Dependent Variable: I’d like you to rate the restaurants you, yourself, have eaten from in the past three months using a 10-point scale, where “10” means you think it is perfect, and “1” means you think it is terrible. F

df1

df2

Sig.

1.584

494

19

.117

Tests the null hypothesis that the error variance of the dependent variable is equal across groups.

a. Design: Intercept + s1 + s2 + d1_3 + d3a_1 + d3a_2 + d3a_3 + d3a_4 + d6 + d5 + d7 + d8 + region

Tests of Between-Subjects Effects

Source

Type III Sum of Squares df

Mean Square F

Sig.

Corrected Model 157.795a

43

3.670

1.220

.167

Intercept

375.637

1

375.637

124.864

.000

s1

25.679

4

6.420

2.134

.076

s2

.051

1

.051

.017

.897

d1_3

10.879

4

2.720

.904

.461

d3a_1

15.880

6

2.647

.880

.509

d3a_2

7.470

3

2.490

.828

.479

d3a_3

5.876

3

1.959

.651

.583

d3a_4

10.036

2

5.018

1.668

.190

d6

23.141

6

3.857

1.282

.264

d5

21.707

4

5.427

1.804

.127

d7

21.387

5

4.277

1.422

.215

d8

.042

1

.042

.014

.906

region

10.354

3

3.451

1.147

.330

Error

1413.932

470

3.008

Total

29606.000

514

Corrected Total

1571.728

513

a. R Squared = .100 (Adjusted R Squared = .018) As we can see, the model itself is not significant. All the variables have significance values greater than 0.05. Hence we can say that rating q9_1 does not differ for various demographic characteristics such as the variables mentioned above.

Variable q9_7 An N-way anova has been done by taking variable q9_7 (Burger king’s rating) as the dependent variable and s1, s2, d1, d3a, d5, d6, d7, d8, region as dependent variables. The following results were observed:

. Source


Mean Square F

Sig.


42

4.628

1.393

.053

Intercept

468.086

1

468.086

140.853

.000

s1

17.127

4

4.282

1.288

.273

s2

6.823

1

6.823

2.053

.152

d1_3

9.300

4

2.325

.700

.592

d3a_1

40.369

6

6.728

2.025

.060

d3a_3

10.291

3

3.430

1.032

.378

d3a_2

10.809

4

2.702

.813

.517

d3a_4

26.800

3

8.933

2.688

.045

d5

4.772

4

1.193

.359

.838

d6

40.431

6

6.739

2.028

.060

d7

13.780

5

2.756

.829

.529

d8

2.841

1

2.841

.855

.355

Error

2618.703

788

3.323

Total

42641.000

831

Corrected Total

2813.071

830

a. R Squared = .069 (Adjusted R Squared = .019)

As we can see, the model itself is not significant. Only the variable d3a_4 (number of children between age 12-17 live in a home) is significant and all other variables are insignificant. So, the restaurant rating q9_7 only differs for variable d3a_4 Q2. Do the four groups defined by “ the extent to which you find it difficult to make up your mind about which fast food restaurant to go to” (q13) differ in their restaurant ratings (q9_1,q9_7,q9_26,q9_36,q9_39)?

Variable q9_1 We will take q9_1 as the dependent variable and q13 as the independent variable and run a one-way anova. The results are as follows

Tests of Between-Subjects Effects Dependent Variable: I’d like you to rate the restaurants you, yourself, have eaten from in the past three months using a 10-point scale, where “10” means you think it is perfect, and “1” means you think it is terrible.

Source


Mean Square

F

Sig.

Corrected Model

27.005a

3

9.002

2.980

.031

Intercept

5586.326

1

5586.326

1.849E3

.000

q13

27.005

3

9.002

2.980

.031

Error

1706.598

565

3.021

Total

32824.000

569

Corrected Total

1733.603

568

a. R Squared = .016 (Adjusted R Squared = .010) As we can see, the model is significant. Also, the variable q13 is significant. So, there is difference among the four groups of q1 w.r.t the restaurant ratings q9_1 (Arby’s ratings)

Variable q9_7 We will take q9_7 as the dependent variable and q13 as the independent variable and run a one-way anova. The results are as follows Tests of Between-Subjects Effects Dependent Variable: I’d like you to rate the restaurants you, yourself, have eaten from in the past three months using a 10-point scale, where “10” means you think it is perfect, and “1” means you think it is terrible.

Source

Type III Sum of Squares Df

Mean Square

F

Sig.

Corrected Model

41.886a

3

13.962

4.139

.006

Intercept

8811.063

1

8811.063

2.612E3

.000

q13

41.886

3

13.962

4.139

.006

Error

3116.975

924

3.373

Total

47283.000

928


Source

Type III Sum of Squares Df

Mean Square

F

Sig.

Corrected Model

27.005a

3

9.002

2.980

.031

Intercept

5586.326

1

5586.326

1.849E3

.000

q13

27.005

3

9.002

2.980

.031

Error

1706.598

565

3.021

Total

32824.000

569

Corrected Total

1733.603

568

a. R Squared = .016 (Adjusted R Squared = .010) As we can see, the model is significant. Also, the variable q13 is significant. So, there is difference among the four groups of q1 w.r.t the restaurant ratings q9_1 (Arby’s ratings)


Tests of Between-Subjects Effects Dependent Variable: I’d like you to rate the restaurants you, yourself, have eaten from in the past three months using a 10point scale, where “10” means you think it is perfect, and “1” means you think it is terrible.

Source


Mean Square

F

Sig.

Corrected Model

41.886a

3

13.962

4.139

.006

8811.063

1

8811.063

2.612E .000 3

q13

41.886

3

13.962

4.139

Error

3116.975

924 3.373

Total

47283.000 928

Corrected Total

3158.861

Intercept

.006

927

a. R Squared = .013 (Adjusted R Squared = .010) As we can see, the model is significant. Also, the variable q13 is significant. So, there is difference among the four groups of q1 w.r.t the restaurant ratings q9_7 (Burger King’s ratings)




Source

Mean Square F

Sig.


3

16.192

3.921

.008

Intercept

8916.302

1

8916.302

2.159E3

.000

q13

48.575

3

16.192

3.921

.008

Error

4703.850

1139

4.130

Total

54241.000

1143

Corrected Total

4752.425

1142

a. R Squared = .010 (Adjusted R Squared = .008) As we can see, the model is significant. Also, the variable q13 is significant. So, there is difference among the four groups of q1 w.r.t the restaurant ratings q9_26 (Mc Donald’s ratings) Variable q_36 We will take q9_36 as the dependent variable and q13 as the independent variable and run a one-way anova. The results are as follows Tests of Between-Subjects Effects Dependent Variable: I’d like you to rate the restaurants you, yourself, have eaten from in the past three months using a 10-point scale, where “10” means you think it is perfect, and “1” means you think it is terrible. Source


Mean Square F

Sig.


3

6.457

2.167

.090

Intercept

9908.298

1

9908.298

3.325E3

.000

q13

19.372

3

6.457

2.167

.090

Error

2720.689

913

2.980

Total

58225.000

917

Corrected Total

2740.061

916

a. R Squared = .007 (Adjusted R Squared = .004) As we can see, the model is not significant. Also, the variable q13 is not significant. So, there is no difference among the four groups of q1 w.r.t the restaurant ratings q9_36 (Subway’s ratings)

Variable q_39 We will take q9_39 as the dependent variable and q13 as the independent variable and run a one-way anova. The results are as follows


Source


Mean Square

F

Sig.

Corrected Model

55.879a

3

18.626

6.463

.000

Intercept

11034.666

1

11034.666

3.829E3

.000

q13

55.879

3

18.626

6.463

.000

Error

2717.500

943

2.882

Total

56893.000

947

Corrected Total

2773.379

946

a. R Squared = .020 (Adjusted R Squared = .017) As we can see, the model is significant. Also, the variable q13 is significant. So, there is difference among the four groups of q1 w.r.t the restaurant ratings q9_39 (Wendy’s ratings)

Chapter 17: Q1. Can each of the restaurant ratings (q9_1, q9_7, q9_26, q9_36, q9_39) be explained in terms of the ratings on the psychographic statement (q14_1, q14_2, q14_3, q14_4, q14_5, q14_6 and q14_7) when the statements are considered simultaneously?

Answer: Arby’s

As we can see that all the significance value of all the variables is greater than .05. So we can say that the ARBY’S rating cannot be explained by any of the 7 psychographic statements.

Burger King

From the given table we can see that significance value of 3 psychographic statements is less than .05. Hence the restaurant ratings can be explained by these statements. Burger King rating is affected by the following variables:  I consider the amount of fat in the foods I eat at fast food restaurants  I have been making an effort to look for fast food choices that have better nutritional value than the foods I have chosen in the past  I am eating at fast food restaurants less often out of concern for the high fat content in the foods at fast food restaurants

McDonald’s

From the given table we can see that significance value of 1 psychographic statement is less than .05, hence McDonald’s can be explained by the following variable:  I am eating at fast food restaurants less often out of concern for the high fat content in the foods at fast food restaurants

Subway

From the given table we can see that significance value of 1 psychographic statement is less than .05, hence Subway can be explained by the following variable:  I am eating at fast food restaurants less often out of concern for the high fat content in the foods at fast food restaurants

Wendy’s

As we can see that all the significance value of all the variables is greater than .05. So we can say that the WENDY’s rating cannot be explained by any of the 7 psychographic statements.

Chapter 18 Q1. Can the males and females (S2) be differentiated based on the ratings on the psychographics statements (q14_3, q14_4, q14_5, q14_6, and q14_7) when the ratings are considered simultaneously? Run a two group discriminant analysis. Then run a logit analysis. Compare the results from the two analyses. 1) Discriminant Analysis Tests of Equality of Group Means Wilks' Lambda F

df1

df2

Sig.

I try to stay current on the latest health and nutrition .989 information

9.734

1

874

.002

I read nutritional labels on .987 most products I buy

11.549

1

874

.001

I am making more of an effort to find out about the nutritional content of the .989 foods I eat at fast food restaurants

9.577

1

874

.002

I consider the amount of fat in the foods I eat at fast .987 food restaurants

11.642

1

874

.001

I consider the amount of fat in the foods my kids .997 eat at fast food restaurants

2.385

1

874

.123

I have been making an effort to look for fast food choices that have better .987 nutritional value than the foods I have chosen in the past

11.959

1

874

.001

Tests of Equality of Group Means Wilks' Lambda F

df1

df2

Sig.

I try to stay current on the latest health and nutrition .989 information

9.734

1

874

.002

I read nutritional labels on .987 most products I buy

11.549

1

874

.001

I am making more of an effort to find out about the nutritional content of the .989 foods I eat at fast food restaurants

9.577

1

874

.002

I consider the amount of fat in the foods I eat at fast .987 food restaurants

11.642

1

874

.001

I consider the amount of fat in the foods my kids .997 eat at fast food restaurants

2.385

1

874

.123

I have been making an effort to look for fast food choices that have better .987 nutritional value than the foods I have chosen in the past

11.959

1

874

.001

I am eating at fast food restaurants less often out of concern for the high fat .994 content in the foods at fast food restaurants

4.937

1

874

.027

Here we can see that the question or variable “I consider the amount of fat in the foods I my kids eat at fast food restaurant” is not statistically significant. So, there is no difference between males and females w.r.t this question.

Test Results Box's M

24.894

F

Approx.

.882

df1

28

df2

2.580E6

Sig.

.645

Tests null hypothesis of equal population covariance matrices.

Box’ test has been accepted. So the covariance matrix for males and females are the same.

Wilks' Lambda

Test of Function (s) Wilks' Lambda 1

.978

Chi-square 19.132

df 7

Sig. .008

Here, the wilk’s lambda is significant. So the mean of discriminant function is different for males and females. The model is able to differentiate between males and females w.r.t the variables.

Standardized Canonical Discriminant Function Coefficients Function 1 I try to stay current on the latest health and nutrition information

.095

I read nutritional labels on most products I buy

.386

I am making more of an effort to find out about the nutritional content of -.041 the foods I eat at fast food restaurants I consider the amount of fat in the foods I eat at fast food restaurants

.648

I consider the amount of fat in the foods my kids eat at fast food -.688 restaurants I have been making an effort to look for fast food choices that have .637 better nutritional value than the foods I have chosen in the past I am eating at fast food restaurants less often out of concern for the high -.192 fat content in the foods at fast food restaurants

Structure Matrix Function 1 I have been making an effort to look for fast food choices that have better .785 nutritional value than the foods I have chosen in the past I consider the amount of fat in the foods I eat at fast food restaurants

.774

I read nutritional labels on most products I buy

.771

I try to stay current on the latest health and nutrition information

.708

I am making more of an effort to find out about the nutritional content of .702 the foods I eat at fast food restaurants I am eating at fast food restaurants less often out of concern for the high .504 fat content in the foods at fast food restaurants I consider the amount of fat in the foods my kids eat at fast food .350 restaurants

If we see the standardized canonical discriminant function coefficient and the structure matrix, variable q14_6 which has high discriminant loading has a lower standardized discriminant function coefficient. This may be due to multicollinearity.

Classification Resultsa

Original

Count %

Are you…?

Predicted Group Membership Male

Female

Total

Male

227

184

411

Female

208

257

465

Male

55.2

44.8

100.0

Female

44.7

55.3

100.0

a. 55.3% of original grouped cases correctly classified. Finally if we see the classification matrix, the hit ratio is only 55.3% which is very low. Although this model is classifying the respondents, it is not able to do the classification accurately. The validity of the model is very low. So, the males and females cannot be differentiated accurately.

2) Logistic regression Again, a logistic regression is ran using the same variables to predict whether the respondent is male or female. The result and the interpretation is as follows

Model Summary Step

-2 Log likelihood

Cox & Snell R Square

Nagelkerke R Square

1

1191.828a

.022

.029

a. Estimation terminated at iteration number 3 because parameter estimates changed by less than .001.

As we can see the Cox and Snell R square and Nagelkerke R square are very low. This shows that the independent variables are not able to predict the dependent variables effectively.

Classification Tablea Predicted Are you…? Male

Female

Percentage Correct

Male

142

269

34.5

Female

114

351

75.5

Observed Step 1

Are you…?

Overall Percentage

56.3

a. The cut value is .500 Here, the overall percentage correct is 56.3 which is very low. The validation is not satisfactory. Prediction accuracy is very low.

Variables in the Equation Step 1a

B

S.E.

Wald

df

Sig.

Exp(B)

q14_1

.024

.097

.063

1

.802

1.025

q14_2

.091

.089

1.038

1

.308

1.095

q14_3

-.010

.103

.009

1

.923

.990

q14_4

.149

.099

2.263

1

.132

1.161

q14_5

-.161

.084

3.681

1

.055

.851

q14_6

.151

.097

2.426

1

.119

1.163

q14_7

-.046

.078

.341

1

.559

.955

Constant

-.529

.223

5.646

1

.017

.589

a. Variable(s) entered on step 1: q14_1, q14_2, q14_3, q14_4, q14_5, q14_6, q14_7.

Here, we can see that all the variables are not significant w.r.t at 5 % significance level. This explains why the model is not good. Therefore, the model is not able to distinguish between males and females w.r.t the given variables. To compare the two-group discriminant analysis and logistic regression, we can say that logistic regression is better at saying that the model is not good than the two-group discriminant analysis. The two- group discriminant analysis show that the variables have discriminating power but the accuracy of model is not good, while the logistic regression outrightly rejects the model. So, logistic regression is better than discriminant analysis in this case.

Chapter 19 Q1. Factor analyze the psychographic statement (q14_1 to q14_7). Use principal component analysis with varimax rotation. Interpret the factors. Answer: Upon using factor analysis for the data set with the variables q14_1 through q14_7 as variables to constitute factors, one factor is formed. Appropriateness of Factor Analysis:

The KMO value of 0.922 indicates that factor analysis. Determination of number of factors:

Based on the Eigen value interpretation, there is only factor formed and variable 1 i.e., q14_1 explains maximum percentage of the factor, by 71%. Rotation of factors:

Since only one factor is formed, rotation was not performed in spite of selecting the option for varimax rotation while performing factor analysis. Interpretation of Factors:

Since only one factor was formed out of all the variables, the factor loading plot was not plotted. Then the interpretation is restricted to the Eigen values in the Principal Component Analysis table, the calculation of factor scores from the coefficient matrix, the scree plot and determining the model fit using the residuals between observed and computed correlation values.

The scree plot indicates the presence of a single factor.

The eigen values indicate that 71% of the variance is cumulatively explained by the factor composed of all the variables considered.

The factor score can be computed using the above table. Factor1 = 0.170 (q14_1) + 0.164 (q14_2) + .... + 0.156 (q14_7).

Goodness of fit of the model:

As per the residual computation between the observed and computed correlations, there were 10 residuals with a value greater than 0.05 which does not indicate a good fit of the model. There might be a requirement of reconsidering the model. The above results indicate that all the health conscious factors can be singly accommodated into one factor. All the variables considered for such analysis indicate the health consciousness of a customer. Hence, while modeling customer preferences based on the health aspects, other variables should also be considered. Such consideration will prove beneficial while offering new products or while assessing how existent products will fit into the new market and its customers.

Chapter 20 Q1. How would you cluster the respondents based on the psychographic statement (q14_1,q14_2,q14_3,q14_4,q14_5,q14_6, andq14_7)? Interpret the resulting clusters. Complete Linkage(Farthest Neighbour) Agglomeration Schedule Stage

Cluster Combined Cluster 1 Cluster 2

Coefficients

Stage Cluster First Appears Next Stage Cluster 1 Cluster 2

1 2 3 4 5 6

1 3 3 3 3 1

716.000 733.000 823.000 999.000 1239.000 1396.000

0 0 2 3 4 1

2 4 6 5 7 3

0 0 0 0 0 5

6 3 4 5 6 0

A cluster analysis was run on 1450 respondents, each responding to items on psychographic statement questionnaire on Wendy’s Fast Food Chain on their attitude towards healthy eating at Fast Food Restaurants. We used a hierarchical clustering analysis method in which we used the Furthest Neighbor method in which the intervals were measured using Sq. Euclidean Distances. From the above Fig 1 and Table 1 we can see that 4 clusters were formed using the dendogram. Q14_1 & Q14_2: Cluster 1 Q14_3 & Q14_4: Cluster 2 Q14_6 & Cluster 2: Cluster 3

Q14_5 & Cluster 3: Cluster 4 Cluster 1: People who are keen on being updated regarding latest information about nutrition and health. Cluster 2: Make an effort on knowing the nutrition value of the food before consumption. Cluster 3: Loyal towards foods with better nutrition value. Cluster 4: People keeping a check on the nutritional content of the food consumed at fast food restaurants and self.

Chapter 21 Q1. Provide similarity ratings on a scale of 1 to 7 for all possible pairs of the following brands of fsast-food restaurants: Arby’s, Burger King, Church’s, Domino’s Pizza, KFC, McDonald’s, Pizza Hut, Subway, Taco Bell and Wendy’s. Develop a two-dimensional MDS map. Interpret the dimensions and the map. Answer: The dataset for the exercise to perform an MDS was constructed based on direct approach of perception data. The restaurants were measured on a likert scale of 1 to 7 based on similarity judgements.

A stress level of 0.03 indicates the model has good to excellent fitness of data.

From the perceptual map above, dimension 1 can be interpreted as price and dimension 2 to be popularity of the brand. As per the map, the QSRs Wendy’s KFC, McDonalds, Church’s and Burger King are low price restaurants which are quite highly popular and they cluster together in the map. The restaurants Pizza Hut and Domino’s Pizza are also high on popularity whereas they are perceived to be quite high on the expenses, as per the map. Due to the similarity in the offerings, these restaurants also cluster together on the map. Then the low on popularity but medium price range restaurants Arby’s, Subway and Taco Bell cluster together on the lower side of the map.

Chapter 23 Q1. Write a report for Wendy’s management summarizing the results of your analyses. What recommendations do you have for the management?

Answer: Defining the Management Decision Problem: As per the analyses and research, the following recommendations are made to the Wendy’s fast food chain management. The Management Decision Problem formulated as per the scenario presented was – “Which geographic locations in the United States should Wendy’s expand in?” Based on this management decision problem, the following Market Research Problem was arrived at – “Determine the customer preference towards consumption of fast food in the determined location with respect to the competitors and the manner in which they would like to be served”. An Appropriate Research Design that would help the management make this decision is presented in the chart below. As per the depiction, the hypothesis would be that expansion will help catalyze growth of the company. Market research will be conducted with the methods of qualitative research by employing the methods of questionnaire, secondary data analysis, focus group interviews and fieldwork. The objectives to be achieved through this exercise are to find out the perception about Wendy’s as a fast food chain, the existence of demand and the ROI of venturing into a new market and how it will help the company grow.

Competitor Analysis: Analysing the market shares of major fast food chains in the US gives an insight into how big the market of fast food restaurant is and which are the major competitors. The following chart illustrates this information.

Mkt.Share in % of Major Fast Food Chains in U.S in 2013 Mc Donald's Corporation

21.7

Yum Brands Inc

8.1

Doctor's Associates Inc.

6.7

Wendy's International Inc.

5.5

Burger King Corporation

4.1

Other

53.9 0

10

20

30

40

50

60

Mkt.Share in %

McDonald’s by far held the largest market share of the United States fast food industry in 2013. Its closest competitor was Yum Brands - owner of popular chains Taco Bell, KFC, Pizza Hut and Wing Street. The top five brands accounted for just under half of the entire U.S. fast food industry which in 2013 generated over 191 billion U.S. dollars in revenue. This revenue was forecasted to rise above 210 billion dollars in 2018. An analysis of the competitors gives an insight into how big market players they are, how deep the pockets of such established businesses are and how established they are. This analysis helps Wendy’s prepare for the next big step of entering a foreign land. So the biggest player that Wendy’s would have to counter in the US market will be McDonald’s.

New Product Development for the foreign market: Wendy’s tried developing a new product, a fish sandwich with Cajun Taste. In order to determine the consumer preference for Wendy’s newly developed fish sandwich mall intercept form of personal interview survey can be used.

Given below are the fact and findings from various researches to substantiate the importance and advantages of using mall-intercept survey method: Mall intercept surveys are widely used and (theoretically) able to reach a large segment of the population. In any given two-week period about 2/3 of U.S. households shop one or more times at a mall. According to a CASRO membership survey, about 25% of all marketing research and 64% of personal interviews are conducted at malls. The advantages of mall samples are: 1) Experimental control. 2) Ability to see things. 3) Availability of kitchens, etc. 4) Minimal Cost. Effect of Mall Samples on Results: 1) For copy, concept, and product tests, data suggest that mall samples understate scores. 2) Ossip reported four studies that found lower top box concept scores for mall surveys compared to door to door, even after controlling demographic differences. 3) Gannon reported study comparing mall and mail panel for a concept/product test. Mall study got lower concept top box but higher product test attribute ratings. How to select an "Ideal" Mall Sampling Plan? 1) Randomly select states or regions. 2) Randomly select cities within region. 3) Randomly select malls within cities. 4) Post interviewers at randomly selected mall entrances. 5) Interview all days and all times mall open.

6) Count traffic so interviews are proportional to traffic based on day of week and time of day. 7) Determine frequency of mall shopping and weight sample so that frequent shoppers not overrepresented.

Advertising expenditures: Wendy’s is the third hamburger chain by sales after McDonalds and Burger King. Although having a major market share, it needs to brand itself in such a way that people are more drawn towards its quality and optimal price. Wendy’s has introduced various new meals in their list which emphasizes on higher quality, great taste and fresh and never frozen ground beef. So in order to study the customer’s awareness of the competitors and how they respond to the new meals, Wendy’s can perform standard test marketing for their new meals by introducing them to customers and collecting reports of what they thought about the new meals. And if the initial findings are found successful, they can expand the same test to different cities and also ask whether they would like to have any changes in the taste or quality or price and act accordingly. It can also help in determining how the consumers rank Wendy’s in comparison with its competitors.

Selection of field workers for the survey: A survey is a method of descriptive research design which is in turn, a conclusive research methodology. The purpose of such a technique would be to arrive at a conclusion so as to address a problem. In order to address the marketing research problem that has been defined, we probe into what a survey actually means. A survey is a structured questionnaire given to a sample of a population and designed to elicit specific information from respondents. The field force is made up of both actual interviewers and supervisors involved in data collection. Since a survey involves less interaction except for interviews, requirement of such personnel is limited. However, there exists a potential for bias in (1) selecting respondents – selecting the incorrect sample (2) asking questions – omitting certain questions (3) recording answers – recording incorrectly or incompletely. Interviewers can influence the bias in their own ways – inflection, tone of voice, suggesting answers, etc.

Hence, while selecting the fieldworkers, care should be exercised to avoid the above mentioned possibilities which might flaw the research of hamper its results. In a computer based or internet survey, such occurrences are low. Hence a team of supervisors must be selected to train them and to supervise the interview process. If interviews are conducted across geographies, the scope of such supervision is limited.

Factor analysis for variables: The results indicate that all the health conscious factors can be singly accommodated into one factor. All the variables considered for such analysis indicate the health consciousness of a customer. Hence, while modeling customer preferences based on the health aspects, other variables should also be considered. Such consideration will prove beneficial while offering new products or while assessing how existent products will fit into the new market and its customers.

Clusters of Customers: Cluster 1: People who are keen on being updated regarding latest information about nutrition and health. Cluster 2: Make an effort on knowing the nutrition value of the food before consumption. Cluster 3: Loyal towards foods with better nutrition value. Cluster 4: People keeping a check on the nutritional content of the food consumed at fast food restaurants and self.

Interpretation of the perceptual map to aid in market entry decision:

From the perceptual map above, dimension 1 can be interpreted as price and dimension 2 to be popularity of the brand. As per the map, the QSRs Wendy’s KFC, McDonalds, Church’s and Burger King are low price restaurants which are quite highly popular and they cluster together in the map. The restaurants Pizza Hut and Domino’s Pizza are also high on popularity whereas they are perceived to be quite high on the expenses, as per the map. Due to the similarity in the offerings, these restaurants also cluster together on the map. Then the low on popularity but medium price range restaurants Arby’s, Subway and Taco Bell cluster together on the lower side of the map. Wendy’s is also perceived to be similar to the McDonald’s, KFC and other QSR type restaurant. There exists competition in this sector as per the initial competition analysis performed. While entering a market like the USA which has established players and the market is mature, the strategy that Wendy’s has to adopt is to be thought about by the management more with respect to sustainability in the area.

Market Research

Recommend Documents