The demographic data, from the secondary research survey, allows the business manager to target candidates in India that match up with the consumer base in the US and Europe. Although culture and habits may differ in India, the selection of cities with a western-type and more modern culture improves the odds of matching consumers to Starbucks premium product line. Leveraging the set of leisure and lifestyle data enables the business manager to judge the receptivity of consumers to Starbucks mall-based sales outlets. The final key piece of secondary research data relates to the current competitors in the coffee and premium coffee market segments. The number of coffee outlets provides insight into both the strength of the competition and the receptivity of Indian consumers to coffee products. Starbucks primary competitor in India is the Barista Coffee Company.
Organizing and comparing all the secondary research data along with the population and per capita information provides the consumer critical economic and market segment data needed for crafting effective sales and marketing strategies for potential coffee consumers in India. Data analyses indicate that Starbucks should open premium coffee outlets in Mumbai and Delhi, future expansion will depend upon the marketing successes in affluent areas with larger population bases, The company should be cognizant of the impact to their profit and revenue streams, assuming they are favorable, they should consider opening several outlets in these high-potential cities to maximize exposure and minimize logistical expenses.
Starbucks has placed several constraints on the research efforts that could affect the organization's decision to enter the global marketplace; however, the selection of India, resources, including budget constraints, language and other cultural barriers, and secondary research are the most critical. Each of these constraints must be addressed and a plan of action must be devised so that these obstacles may be overcome. Limiting research efforts will affect the overall outcome of the findings, and therefore, limiting the possibilities of success. Identifying limitations is one thing, but finding a solution should be the primary focus.
The below SWOT Analysis was taken from Datamonitor, 2007 and highlights the challenges as well as the opportunities for Starbuck's within the Indian
Markets:
Strengths
1. Strong brand image
2. Robust financial performance
3. Large scale of operations
Weaknesses
1. Weak compliance function
2. Narrow product mix
3. Low employee productivity
Opportunities
1. New markets
2. Hear Music Label
3. Growing specialty coffee market
Threats
1. Intense competition
2. Volatile coffee prices
3. The Forbidden City controversy
One of the main areas Starbucks can benefit from is robust financial performance. Tapping into growing markets within growing countries with robust financial performance and opportunities can be very financially rewarding and should be strongly considered.
When organizations consider expanding into countries other than its country of origin, language and cultural barriers can pose potential problems. Although Hindi is the national language, 24 other languages exist within the country. English is also an important language for political and business purposes within the country. Furthermore, 80% of the population is Hindu, a religion not commonly practiced within the United States. Effective communication is vital to research success. Although English is spoken, the researcher must understand and respect the cultures of the Indian people.
Hypotheses and Testing
Null Hypothesis
The Starbucks Coffee Corporation seeks to validate or invalidate a hypothesis associated with expanding operations into in one or more cities in Austarlia. Starbuck's marketing team should craft a null hypothesis, H0, addressing the potential that the Indian culture, infrastructure and economics do not support premium coffee sales. The following constitutes a recommended null hypothesis tied to the potential expansion into the Indian coffee market:
• The Indian culture, infrastructure and economy will not support future premium coffee sales at target market areas within India.
The market research team intends to test Starbucks' null hypothesis, statistically, to show that no difference exists within the groups tested or that no correlation exists between the variables i.e. that zero potential exists from premium coffee sales in target areas. If, however, data indicates that the null hypothesis can be rejected, the existence of a difference or a relationship can be proved at the desired confidence level then the company may decide to move forward with global expansion plans (Wallace, & Perlman, 1979).
The research team presumes that statistical testing will validate the null hypothesis. The null hypothesis will be rejected only when statistically proven false, based upon the selected degree of confidence 95%. Fisher, who first coined the term "null hypothesis" stated that, "the null hypothesis must be exact, that is free of vagueness and ambiguity, because it must supply the basis of the 'problem of distribution,' of which the test of significance is the solution" (Fisher, pg. 30, 1948).
Alternative Hypothesis
A null hypothesis supports calculating the probability of observing a data set with particular parameters. Typically, researchers face the challenge of determining with precision how probable the data would be if the alternative hypothesis were true. In the case of Starbucks the market research team does not seek to validate the alternative hypothesis through statistical testing.
Executives at Starbucks wish to disprove the null hypothesis so that the organization can expand operations thus boosting sales revenues and increasing shareholder value. Should the multivariate data set support rejection of the stated null hypothesis the statistical analysis process requires an alternative hypothesis H1, one that supports Starbuck's mission and vision. The recommended alternative hypothesis states that:
• The Indian culture, infrastructure and economy support expanding Starbucks' operations, premium product sales, into targeted regions of India.
The complexities involved with collecting and analyzing financial, infrastructure, demographic and market data present significant challenges for the global market research team. Inadequate or inaccurate data could result in a false failure to reject the null hypothesis leading to a faulty business decision. The financial investment requires a high level of confidence that the survey data and associated analysis accurately represent the true state of conditions within India, at a 95% probability level.
Hypothesis Testing
As part of Starbucks' statistical testing the alternate hypothesis will form the basis of future business decisions if a sample contains sufficient evidence to reject the null hypothesis. The stated alternative hypothesis communicates the expected conclusion of statistical testing, support for market expansion into India. Should the statistical tests fail to disprove the null hypothesis then the research team will fail to reject the null hypothesis rather than stating acceptance. Starbucks' statistical testing will not prove the null hypothesis; instead it will disprove the hypothesis and reject it (Wallace, & Perlman, 1979).
Starbucks' business plan requires the use of multiple regression and Chi-Square tests as methods for validating data and developing projections that fall within the confidence levels specified by Starbucks' market research team. The survey data set includes financial time series data related to historical sales of premium coffee products (including tea) in the Indian market segment (see Appendix F). Team members anticipate that the data will support a forecast of an expanding premium coffee and tea market in India thus enabling Starbucks to capitalize on favorable market conditions.
Starbucks' research team selected Fisher's method, developed by and named for Ronald Fisher, to complete data fusion or meta-analysis, i.e. analysis after analysis, allowing the team to combine the results from a variety of independent tests bearing upon the same overall null hypothesis (H0) as if in a single large test.
Fisher's method combines value probabilities, from sample data into a single test statistic (X2) having a chi-square distribution, refer to the formula listed below (Dickerson & Berlin, 1992).
Data Sources and Sample Size
When making business decisions based on data retrieved through sampling, an appropriate sample size be used, this is critical. The MICS3 Manual (2007) touts that "the size of the sample is perhaps the most important parameter of sample design, because it affects the precision, cost and duration of the survey more than any other factor." Much forethought and consideration is necessary when determining sample size, since surveying every member of India's ever-growing population may not be economically feasible or possible. Equally important as the budget is the level of precision required. According to Lind, et al (2005), if a sample is too small, the conclusions drawn from the sample may be skewed and uncertain. Similarly, if the sample size is too large, money is wasted collecting data.
Therefore, Lind et al (2005) contend that the necessary and appropriate sample size depends on three factors:
1. The level of confidence desired.
2. The margin of error the researcher will tolerate.
3. The variability in the population being studied.
Ideally, sampling the entire population would provide us with answers to all our research questions but that is an impossible task to accomplish. The most effective option is to select a sample size that is large enough that adequately represents the characteristics of the population, yet data collection is cost-effective. Based on the information provided in the CoffeeTime scenario, a sample size of 60 or even 90 would be appropriate. Either of these sample sizes would not only be budget friendly, but a sample size of 60 or 90 would also provide adequate information within a reasonable margin of error.
As Lind, et al (2005), state that the variability in the population being researched must also be considered. "If the population is widely dispersed, a large sample is required. On the other hand, if the population is concentrated (homogenous), the required sample size will be smaller" (Lind, et al, 2005, p. 302). According to the research (see Appendixes D and E), the population within New Delhi and Mumbai is fairly homogenous, especially when affluence is one of the primary factors considered. Based on this information, a sample size of 60 or 90 is adequate and recommended.
Selecting and producing samples is a rigorous task and one that should not be taken lightly. Starbucks must be diligent to ensure the data is valid, accurate and reliable. Examining the respective city's demographics is essential. Starbucks must gather data relative to gender, level of affluence and cultural preferences. Additionally, data supporting monthly income, infrastructure, the number of local restaurants within each respective city and information related to competition is essential. Evidence must exist that concludes Indian citizens will support the premium products that Starbucks offers as well as their growth plan within the Indian Markets. Without this information, Starbucks would be taking a significant financial risk with inconclusive primary and secondary data to support projections of favorable profit and revenue projections.
In order to gain this information, Starbucks will need to gather data through surveys and perhaps other sampling methods. Some survey questions for additional questions/data Starbucks should consider asking are:
a. Do the target groups drink coffee? If so, much per day?
b. What is his or her income bracket?
c. What is their gender?
d. What types of coffee do they enjoy?
e. Where do they purchase coffee?
f. Do they frequent coffee bars?
i. If so, how often?
ii. If so, how much do they spend per visit?
iii. What do they look for in a coffee bar?
Starbucks can also serve as a place for meting and socialization. When Starbucks entered China, consumers were not initially drawn to Starbucks for the coffee, but for the socialization opportunities, according to Christine Day, President of Starbucks Asia Pacific Group. Day (2004) continued to explain during her interview with Parija Bhatnagar of CNN/Money, that like China, "India is a tea-based culture coffee is not a substitute. Starbucks is more of a place to hang out, to eat and drink, to see and be seen." In addition to questions regarding one's coffee preferences, questions regarding one's socialization habits would provide Starbucks with a wealth of information.
Researchers must think outside the box when devising survey questions for the sample(s).
Starbucks, in addition to conducting its own research, may also choose examine data that is already available. Starbucks in strategizing its Asian expansion can follow companies such as McDonald's, Citibank and Motorola are "hustling to tap India's burgeoning number of young big spenders" (Schuman, 2003). In regards of the Indian market Although Barista Coffee Company, Starbucks primary competitor, opened its first café in New Delhi in 2000, the information and research may exist, however, the age of the data could skew the results. It would be in Starbucks best interests to conduct their own research for this business venture.
When selecting who to survey, careful consideration and attention must be paid to all age groups. Typically, pre-teens are neither decision makers, nor do they possess the spending power required for Starbucks products. Therefore, it would not be unreasonable or even impractical to omit them from the sample.
Teens tend to have some spending power and may find Starbucks to be a hip and happening place; however, the prices charged for Starbucks products may be a bit expensive for teens. Starbucks offers a diverse range of products which may suit their varieties of tastes and it would be wise for Starbucks to examine any existing data supporting this contention. If no data exists, it may be in Starbucks' best interests to include teens in their surveys and research.
Young adults are decision-makers and have extensive spending power. With New Delhi and Mumbai being home to many call centers where young and middle-aged Indians are employed. Further, "the spending of these college grads is rising about 12% a year – more than twice the pace of the economy's growth" Michael Schuman reported in Time Magazine. R.K. Shukla, a statistician at the National Council stated in his interview with Michael Schuman of Time, "Income is growing like anything .the future is very rosy in India." These citizens must be included in the sample. Research and data gained from these citizens would be most beneficial to Starbucks.
Determining test statistics is no easy task. In this instance, the shape of the population is unknown, but the number of observations is greater than 30. In situations such as these, Lind, et al (2005) argues that "often we can reason that the population is normal or reasonably close to a normal distribution. Under these conditions, the correct statistical distribution is to replace the standard normal distribution with the t distribution, which is a continuous distribution with many similarities to the standard normal distribution." The t distribution must also be used because the population standard deviation is unknown. The t distribution is flatter and more spread out than the standard normal distribution, but this is because the standard deviation of the t distribution is larger than the standard normal distribution. Since Starbucks can reasonably assume the population distribution is normal, using the t distribution will provide reasonable and accurate data on which to base business decisions.
Strategic Decision
The global market research team recommends expanding Starbucks' operations into the following regions of India:
• New Delhi (multiple retail locations)
• Mumbai (Bombay – multiple retail locations)
• Lucknow
• Chandigarh
The comprehensive statistical analysis package completed on behalf of the global market research team supports expansion into specific target areas while indicating that other cities within India may not prove to be viable expansion candidates. Evaluating a limited set of data always presents the risk of poor business decisions based upon faulty assumptions. Statistical testing cannot prove the viability of Starbucks' alternate hypothesis so company executives must weigh the test results using the wealth of professional experience possessed by the Starbucks' organization.
Conclusion
When faced with challenging business decision one should conduct as much research as possible to aide in the decision making process. "There are three components to any decision: (1) the choices available, or alternatives; (2) the states of nature, which are not under the control of the decision maker; and (3) the payoffs. Statistical decision theory is concerned with determining which decision, from a set of possible alternatives, is optimal for a particular set of conditions" (Lind et al, 2005, pp. 688 – 689). The CoffeeTime scenario was an excellent example of how to use statistics to help make viable business decisions that provide plausible short and long-term solutions that will support overall growth projections that will meet or exceed the company's financial goals.
Developing a good research design is very important and having the ability to use secondary research to target resources and develop primary research is also an extremely valuable tool. In research, the limitations and boundaries that one must consider are varied, such as validity of data, sample size and budget constraints. However, the goal remains the same, which is to minimize the risk in order to maximize the rewards. When making any business decision, conducting research provides the overall business with quantitative and qualitative statistical data which can help a business achieve their overall goals and objectives through the gathering, review, analysis and presentation of valid, reliable and accurate data.
References:
Bhatnagar, P. (2004, Nov. 1). Starbucks: A passage to India. Retrieved January 11, 2008 from news/fortune500/starbucks_india/index.htm
Cooper, Donald R., Schindler, Pamela S. (2003). Business Research Methods 8th ed. [University of Phoenix Custom Edition e-Text]. New York: McGraw-Hill Irwin. Retrieved November 31, 2007, from University of Phoenix, rEsource, MBA210 Managerial Decision Making Web site.
CultureGrams. (2007). Retrieved June 3, 2007, from http://online.culturegrams.com
Datamonitor. (2007, April 4). Starbucks corporation. Retrieved January 14, 2008 from http://www.datamonitor.com/home/
Dickerson, K., and Berlin, J. A. (1992). Meta-Analysis: state of the science. Epidemiologic Reviews 14:154–176. Retrieved January 16, 2008 from http://www.bmj.com/archive/7119/7119ed.htm
Fisher, R. A. (1948). Combining independent tests of significance. American Statistician, . 2, issue 5, page 30. Retrieved January 16, 2008 from http://worldcat.org/wcpa/ow/a8bff44ef56d1653.html
Lind, Douglas A., Marchal, William G., Wathen Samuel A. (2005). Statistical Techniques in Business and Economics 12th ed. [University of Phoenix Custom Edition e-Text]. New York: McGraw-Hill Irwin. Retrieved November 29, 2007, from University of Phoenix, rEsource, MBA210 Managerial Decision Making Web site.
Schuman, M. (2003, Aug. 25). Hey, big spenders. Time. Retrieved January 14, 2007 from http://www.time.com/time/magazine/article/0,9171,1005522-2,00.html
University of Phoenix. (2003). Managing Research Design [Computer Software]. Retrieved November 29, 2007, from University of Phoenix, rEsource, Simulation, MBA510 Managerial Decision Making Web site.
United Nations (2006) India: DEMOGRAPHIC PROFILE. United Nations Economic and Social Commission for Asia and the Pacific. Retrieved January 18, 2008 from pend3.htm
USDA (20005) India: Basic Information. United States Department of Agriculture: Economic Research Service. Retrieved January 18, 2008 from http://www.ers.usda.gov/Briefing/India/Basicinformation.htm
Wallace, D.L., Perlman, M.D. (1979). Methods in probability and statistical inference. Final report, June 15, 1975-June 30, 1979. Chicago University, IL (USA). Department of Statistics. Retrieved January 16, 2008 from ://www.osti.gov/energycitations/product.biblio.jsp?osti_id=5047850
Appendix A (Trade – India)
USDA (2005)
USDA (2005)
Appendix B (Income Growth and Poverty Rates – India)
USDA (2005)
Appendix C (Population Projections – India)
United Nations (2006)
Appendix D (Population Map - India)
Above Map taken from www.mapsofindia,com
Appendix E (Per Capita Map – India)
Above Map taken from www.mapsofindia,com
Appendix F (Coffee Consumption Growth % – India)
Goodness of Fit Test
observed expected O - E (O - E)² / E % of chisq
3.4 3.000 0.400 0.053 0.72
8.84 8.000 0.840 0.088 1.19
27.41 20.000 7.410 2.745 36.92
0.14 1.000 -0.860 0.740 9.95
6.41 5.000 1.410 0.398 5.35
3.17 3.000 0.170 0.010 0.13
5.54 5.000 0.540 0.058 0.78
0.25 1.000 -0.750 0.563 7.56
1.41 1.000 0.410 0.168 2.26
1.58 2.000 -0.420 0.088 1.19
9.24 8.000 1.240 0.192 2.58
1.37 1.000 0.370 0.137 1.84
1.94 2.000 -0.060 0.002 0.02
1.16 1.000 0.160 0.026 0.34
1.18 1.000 0.180 0.032 0.44
0 1.000 -1.000 1.000 13.45
5.83 5.000 0.830 0.138 1.85
1.74 2.000 -0.260 0.034 0.45
1.11 1.000 0.110 0.012 0.16
4.28 3.000 1.280 0.546 7.34
1.3 1.000 0.300 0.090 1.21
0.88 1.000 -0.120 0.014 0.19
0.9 1.000 -0.100 0.010 0.13
2.73 3.000 -0.270 0.024 0.33
2.4 2.000 0.400 0.080 1.08
1.25 1.000 0.250 0.063 0.84
1.25 1.000 0.250 0.063 0.84
1.25 1.000 0.250 0.063 0.84
97.96 89.000 12.960 7.436 100.00
7.44 chi-square
27 df
.9999 p-value
Appendix F (Continued)
ANOVA
One factor ANOVA
Mean n Std. Dev
23,611.3 4 2,998.29 T.N.
14,725.8 4 752.91 Karnataka
4,148.3 4 1,091.92 Kerala
4,760.3 4 954.59 AP
47,245.3 4 5,470.52 South India
3,363.8 4 184.63 Non-South
50,608.8 4 5,469.26 All India
21,209.0 28 19,336.28 Total
ANOVA table
Source SS df MS F p-value
Treatment 9,880,473,857.71 6 1,646,745,642.952 161.14 1.90e-16
Error 214,601,047.25 21 10,219,097.488
Total 10,095,074,904.96 27
Post hoc analysis
Tukey simultaneous comparison t-values (d.f. = 21)
Non-South Kerala AP Karnataka T.N. South India All India
3,363.8 4,148.3 4,760.3 14,725.8 23,611.3 47,245.3 50,608.8
Non-South 3,363.8
Kerala 4,148.3 0.35
AP 4,760.3 0.62 0.27
Karnataka 14,725.8 5.03 4.68 4.41
T.N. 23,611.3 8.96 8.61 8.34 3.93
South India 47,245.3 19.41 19.07 18.80 14.39 10.46
All India 50,608.8 20.90 20.55 20.28 15.87 11.94 1.49
critical values for experimentwise error rate:
0.05 3.26
0.01 4.00
p-values for pairwise t-tests
Non-South Kerala AP Karnataka T.N. South India All India
3,363.8 4,148.3 4,760.3 14,725.8 23,611.3 47,245.3 50,608.8
Non-South 3,363.8
Kerala 4,148.3 .7320
AP 4,760.3 .5433 .7892
Karnataka 14,725.8 .0001 .0001 .0002
T.N. 23,611.3 1.28E-08 2.48E-08 4.20E-08 .0008
South India 47,245.3 6.80E-15 9.74E-15 1.29E-14 2.40E-12 8.83E-10
All India 50,608.8 1.55E-15 2.17E-15 2.83E-15 3.59E-13 7.94E-11 .1516
Appendix F (Continued)