Marketing Research Report

Marketing Research Report
This is a challenging project. You are going to conduct a market research project for a real company and then make recommendations to top management based on data you collect. To make it easier for you, I?ve put together this guide to give you a step-by-step outline of what you have to do. Each of the 7 steps below has a dedicated section that will explain how to complete the task.
1. Define the decision
2. Create a conceptual model
3. Design the survey instrument
4. Choose your sample
5. Collect data
6. Analyze data
7. Report your results
8. Evaluating a market research report
1. Define the decision

The topic should be a current marketing dilemma, one where a marketing manager has a decision to make, but it isn?t obvious what the best decision is. Resist the temptation to run a survey that asks a bunch of questions that merely seem relevant; that?s a sure way to collect a lot of nonsense information. The best topics usually fulfill the following:
? Managers are currently facing a dilemma where the best course of action is not clear. If the answer to your question is already obvious, then your research will not be worthwhile. I?ve had people tell me they want to ask people where (which retailer) they want to purchase their product in, when the contract has already been signed to sell in Best Buy or another retailer. If the decision has already been made, then you are wasting your time and money on such a question.
? The decision cannot be made by using other forms of data. If the decision is best made by analyzing sales figures or by asking people in the R&D department then you don?t need to conduct a survey. Surveys are most helpful when you need to know customer preferences, attitudes, usage or products and services, and how your company?s offering might fit into the customer?s life.
? Respondents have opinions about the topic. If people don?t have any opinions or are not interested at all in the topic, you will have pretty boring results as well. Boring in, boring out as they say? Market research works best when respondents are engaged in the topic enough to provide the data you need. The same goes for highly technical questions. As a respondent, I wouldn?t be able to provide any help if you are trying to decide between using a ?variable bypass system? or a ?trans-inductive capacitor? (I just made those up!) in a microwave. I can however, give an opinion on the design, features and performance aspects of the microwave.
Overall, the most researchable decisions are binary decisions (i.e., dilemmas). These are either-or decisions between two different but potentially attractive alternatives. Often, these involve severe trade-offs. Examples include: Should I sell my product in a high-end retailer or a mainstream outlet? Should I target teens or ?twenty-somethings?? If there isn?t a choice, there isn?t anything to test; and if you?re not testing anything, you?re just collecting information for fun.
Your decision should be turned into a precise research question. It might sound something like, ?Should we raise or lower the price of our consulting services?? or, ?Should the packaging on our bags of spinach be blue or green?? A poor research question is very general, without precisely defining the potential decision. This sort of research question often sounds something like, ?What should I do to sell more pancake mix?? or, ?What do our custom chemicals customers want?? It should be clear that the former questions provide clear direction, while the latter two questions don?t.
2. Create a conceptual model

Once you have a clear decision to make, it?s time to develop a conceptual model. A conceptual model explains how you think your decision will affect people?s perceptions and behaviors. It typically involves an illustration ?using boxes and arrows. For this project, and for most projects you will encounter in your career, I recommend using a conceptual model with 4 boxes (see Figure). A conceptual model may involve any number of boxes, but four is best because it provides all the essentials in a way that?s easy to understand. Label the four boxes: Decisions, Perceptions, Behaviors, and Categories.
? Decision(s): This box represents your decision from the previous step. For example, say you sell bathroom cleanser and your decision is whether or not you should claim that it uses bio-degradable chemicals. In this case, write ?claim bio-degradable or not? in the Decision box.
? Perceptions: The second box lists all the things that people might think as a result of your marketing decision. You can also think of perceptions as the reasons people give for their behaviors. You cannot directly observe these thoughts and feelings; they all occur in the person?s mind. However, we can try to discover what these perceptions are. In the example of the bio-degradable chemicals, ask yourself, if you put this on the label what might people think? If you put on the label, they might think, ?this is an environmentally responsible company,? but they might also think, ?the cleaning agent might not be as strong as others.? List both of these thoughts/perceptions in the perceptions box.
? Behaviors: These are observable behaviors that a person might do in response to your decision. Ask yourself, ?if I make the right decision, how will I know based on each person?s change in behavior?? In the example of the bio-degradable cleanser, behaviors might include purchase, usage frequency, and word of mouth to others. List these in the behavior box, they will be turned into questions in the next step.
? Categories: The final box represents types of people (i.e., segments) who would respond differently based on which decision you make. Ask yourself, ?are there types of people who would react especially strongly (either positively or negatively) to my decision?? In our continuing example, you might list ?environmentally conscious? since this group will be especially favorable to the bio-degradable appeal. If younger consumers would be especially responsive to an environmental message, then include age as a category and list the various categories that are important to you.

Example of a Basic Conceptual Model
3. Design the survey instrument

The survey instrument is the questionnaire that you will give to your respondents on paper, by phone, over the web, or through some other means. It includes all instructions, questions, and responses that a respondent needs to complete your survey.
Designing a useful survey instrument requires a combination of a well-defined research question, attention to detail, common sense, and experience. There are technical aspects, but it is also a bit of an art. Above all, always try to make things easy and clear for the person taking your survey. Strive for clarity and ease of use. Your respondent will appreciate it and you?ll get better results.
For this assignment, I recommend having about 3 behavior questions, 3 perceptions questions, and 3 category questions. Having around 10 questions will force you to ask only the most important questions and will keep your analysis more manageable. Most surveys start with behavior questions, then go to perceptions, and finally ask categorical questions. This depends a bit on your research questions, however, so don?t feel that this order is written in stone.
Behaviors
A typical survey will start with behavioral questions. These need to be forward looking, asking about the intent of the respondent to do something in the near-future. Here are some example questions:
 How likely are you to purchase Progresso the next time you buy canned soup?
Extremely Unlikely 1 2 3 4 5 6 7 Extremely Likely
 How likely are you to recommend Verizon Wireless to people you know?
Extremely Unlikely 1 2 3 4 5 6 7 Extremely Likely

Perceptions
The next set of questions should relate to perceptions that drive the behaviors above. Think of this in terms of a respondent explaining, ?I will be very likely to buy Progresso because _____.? Sometimes the perception will appear to be more rational (?Kroger has the best prices?) and sometimes it may be more emotional (?Twittering makes me feel like I?m part of a larger community?). The most important thing is that the perception is something that exists in the mind of the respondent, and that you think it causes one or more of the behaviors above.
 To what extent do you agree with the following statements?
Wal-Mart is ethical.
Disagree strongly 1 2 3 4 5 6 7 Agree strongly
Whenever Wal-Mart enters a city, local merchants suffer.
Disagree strongly 1 2 3 4 5 6 7 Agree strongly

 How well do the following characteristics describe Sony VAIO notebook computers?
High quality Does not describe at all 1 2 3 4 5 6 7 Describes perfectly
High performance Does not describe at all 1 2 3 4 5 6 7 Describes perfectly
Sleek design Does not describe at all 1 2 3 4 5 6 7 Describes perfectly
Long battery life Does not describe at all 1 2 3 4 5 6 7 Describes perfectly
Categories
These questions are designed to compare different types of people?s reactions to marketing actions. In essence, you are testing various segmentation schemes. The best way to come up with these categories is to ask yourself, ?What type of person would react very differently if I make the decision one way versus the other?? For example, what type of person would react most/least strongly to a decrease in price? Males? People aged 50+? Residents of California? If you have a hunch that there will be a difference, create a question so that you can record which group each of your respondents is in. If you don?t think there will be a difference, then don?t bother including a question.
 Which of the following most closely describes your work status? (check one only)
? Student ? Employed full-time ? Employed part-time ? Retired
 Age: __________

 Gender
? Male ? Female ? Prefer not to answer

People sometimes get uncomfortable when others (you, in this case!) prejudge them, putting them into distinct, predetermined categories. For example, many people do not like to be forced into racial categories, preferring to describe themselves as members of multiple ethnic groups. Put yourself in the position of various types of respondents, and change any questions that you think would pigeonhole respondents into inaccurate groups. The US Census encountered this problem when they used to ask people to classify themselves as having a single race/ethnic origin. The Census now allows respondents to choose more than one category, and has made other changes so that today?s data accurately portrays people?s heritage.
Also, take care in asking questions to make sure they are not too personal. Imagine how you would feel if, upon being introduced to someone, they immediately asked you your salary? If you must ask questions about sensitive topics like income, it is often a good idea to create broad categories of income (e.g., ? $0-$34,999 ? 35,000-$49,999 ? $50,000+ ? Prefer not to answer ). This reduces pressure and is more likely to yield an accurate response. Also, for any question that may be perceived as intrusive, include a ?prefer not to answer? category. The bottom line is, don?t make your respondents uncomfortable; it?s unethical and will make your survey less successful.
4. Choose your sample

Surveys are a way to ask questions to a sample of people that you think represent some larger segment of the population. As an analogy, if you want to test the chlorine content in a pool you don?t empty the whole pool and test every drop, you just take a sample of water from the middle and then assume that the rest of the pool is the same. If there is a dead frog in the pool, you wouldn?t take your sample from that part because it might be contaminated. Similarly, you want to choose respondents that best represent the views and thoughts of the entire pool you want to learn about. Choose your sample of respondents carefully as this will be the first criticism you encounter if people aren?t happy with your results.
Make sure that you have adequate numbers of people in each of your categories in the survey. For example, if you have a Male-Female question, make sure that you get some males and some females to complete your survey so that you can compare how males responded compared to females. Thus, the sample you choose depends on the topic you choose, but also on what questions you ask.
5. Collect data

Collecting data is not particularly difficult, but it demands careful attention and can sometimes be time consuming, even frustrating. You have to keep track of who has completed the survey and who hasn?t. It may involve sending reminders. You may also have to provide some sort of incentive to get respondents to complete your survey. Two common methods of doing this are small monetary incentives (say, $2-$10 each person that completes the survey) or a lottery (say, a random drawing will choose a first, second and third prizes of $100, $50, and $25 gift certificates to a store that is appealing to the respondents). For this assignment, you?ll be surveying friends and family, so you may want to use a third method: guilt!
You might have to deliver some questionnaires by phone and others by email. You should keep in mind that people take surveys differently based on the form in which it is delivered. It is usually safest to deliver them all the same way, however for this particular assignment it?s okay to use multiple methods of data collection.
If you decide to collect data online with services such as Zoomerang, or SurveyMonkey, make sure that you are able to conduct a full analysis with the data they give you. Some of these services will only let you calculate the average response for the whole sample; in other words, they don?t give you the capability of calculating the average response on a question for males versus females unless you pay extra. In these cases, it?s better to type you results into excel or another spreadsheet by hand.
Whenever you tabulate data, you should type it into a spreadsheet so that you can analyze it using statistical software. Each row should represent a different respondent; each column will represent a different variable (i.e., question). Suppose your methodology looked something like this:
? Respondents randomly assigned so that half answer all questions without seeing an ad, and the other half respondents answered the same questions after seeing the ad that you are trying to test.
? Your survey questionnaire has 2 behavioral questions (purchase intent and word of mouth), 2 perception questions (quality and ease of use), and 2 categorical questions (male-female and college degree or no college degree)
In this case, when you enter the data (i.e., people?s responses) into excel or another spreadsheet, it should look something like this:
Respondent Group # (randomly assigned to either 1=saw ad, or 0=did not see ad) Q1: Purchase intent Q2:
Word of Mouth Q3:
?This product is high quality? Q4: ?This product will be easy to use? Q5: 1=Female, 0=Male Q6:
1=College degree, 0=No college degree
A 0 3 4 5 4 0 1
B 1 6 5 6 5 0 0
C 0 2 4 6 3 1 0
D 1 7 7 4 7 1 1
E 0 4 3 5 2 1 1
?n? 1 5 7 5 6 0 0
6. Analyze data

If you have designed your survey carefully analyzing you data should not be a terribly complicated exercise. If, on the other hand, you?ve been sloppy early on, this section can become cumbersome. Here is a step-by-step way to analyze this:
a) For each behavioral and perceptions question, calculate the mean (average) and standard deviation. For each category question, calculate the percentage of respondents in each category.
b) Do step ?a? above again for each category that you have recorded. For example, one of our categories above is male or female?calculate the average on each behavioral and perception question for (1) Females, and then for (2) Males.
c) Compare your perceptions questions to your behavior questions to see if people who are high in certain perceptions are high (or low) on any of your behaviors. Statistical methods such as multiple regression and ANOVA are ideal for this sort of analysis. If you aren?t familiar with these basic statistics methods, see me for some alternative ways to look at the data that are simpler (but also more crude).
Once you have your data analyzed in this way. Pick out the numbers that help you make your decision. Anything that does not help you make your decision is interesting, but not particularly relevant. Try to make comparisons between categories of people that help you understand how people form their perceptions and decide on behaviors. Also, using the data at hand, try to identify categories of people that will most be impacted by your decision.
An Example
Based on the data in the spreadsheet in the previous section, your results would look something like this. Notice that each question has the average response for all respondents and then three sets of ?cross tabs? that represent your categories. One of these sets of categories (A) reflects those who saw the ad versus those who did not see ad (recall that you randomly assigned respondents to one group or the other).
A B C
Respondent All respondents Saw ad Did not see ad Female Male College degree No college degree
Q1: Purchase intent 4.5 6.0 3.0 4.3 4.7 4.3 4.7
Q2: Word of Mouth 5.0 6.3 3.7 4.7 5.3 4.7 5.3
Q3: ?This product is high quality? 5.1 5.0 5.3 5.0 5.3 4.7 5.7
Q4: ?This product will be easy to use? 4.5 6.0 3.0 4.0 5.0 4.7 5.4
Q5: 1=Female, 0=Male 50% Female, 50% Male 50% 50%
Q6: 1=College degree, 0=No college degree 50% College degree, 50% No college degree 50% 50%

First look at the ?All respondents column.? These numbers need to be included, but they aren?t very meaningful on their own. Who?s to say that a 5.1 on Q3 is a good score or not? Sure it?s above the mid-point in our 1-7 scale, but if we had asked, ?This product is GOOD quality? instead of HIGH quality, then the answer might be much different. So the ?all respondents? column becomes a baseline that we use to help us make sense of the other categories. The real insight comes from analyzing the cross-tabs.
Do you see any differences between categories of respondents? Let?s compare the two columns under A (people who saw the ad versus people who didn?t see the ad). It looks like there is a substantial difference between groups in their response to Q1: purchase intent. People that saw the ad want to purchase it more than those who didn?t see the ad. Seeing the ad didn?t seem to affect people?s perceptions of the product as high quality though?there is not much difference (5.0 and 5.3), and if anything, seeing the ad seems to have slightly reduced the perception of quality.
How different should these scores be for us to say that two groups are different? For small samples (less than 50) you might use a rule of thumb that differences must be greater than .4 to be meaningful. Ideally, you should use more sophisticated statistics like regressions or ANOVAs to determine if the differences are statistically significant. But the size of the difference also depends on the question. If you just want to know whether gender is a factor in how people purchase, small differences may suffice. But if you want to know whether it?s a good idea to abandon males to target only females, you may want to see a stark difference so that you don?t end up ignoring a very large group of potential customers.
7. Report your results

Title page
Please include a title page for this report. Also, please do not include identifying information on subsequent pages; this helps me grade more objectively.
Motivation/Research Question (around ? of a page)
This section is designed to give the reader a basic background of why you need to conduct the study. Resist the temptation to include a long-winded story. Stick to the facts that motivate your research and that result in a clear decision that the company (i.e., you) needs to make. Explain how market research will help you make this decision. This section typically concludes with your basic research question. In this report, the research question should be a choice between two clear alternatives (e.g., launch/don?t launch?, raise price/lower price?, distribute at superstores or through boutiques?).
Methodology (around 1 page)
This section should describe in very clear terms what you did to answer your research question. First, you should briefly describe and display a simplified version of your conceptual model. For example, only mention the perceptions, behaviors, and categories which were then turned into questions in your survey.
Second, describe how you performed your research. However, do not include any findings yet. You will answer the following questions:
? Who did you invite to answer the survey?
? Is there any special reason why you chose this group? (e.g., ?wanted to have an even number of males and females?)
? How many people responded to the survey?
? How was the questionnaire structured? (You might say that you began with behavioral questions, then examined perceptions, and finally captured characteristics of the respondents such as age, or educational background. You can refer them to the appendix for the detailed list of questions.)
? Did everyone receive the same questionnaire or did you split the respondents into groups with two different conditions?
? Did you deliver the survey in person, over the phone, by email, or through some other means?
? Any other information about how you conducted the survey that you think might be important.
Think of this section of your report as helping the reader understand how you investigated your research question, and enabling them to replicate the study if they want to.
Findings (around 2 pages)
In this section, you report on any data that you think helps you answer the research question. In other words, if it doesn?t help you make the decision that you?ve defined at the beginning, you probably don?t need to talk about it. Make sure to report all information that is related to the decision, even if it contradicts what you think is best. So be honest about what the data is saying.
For the behaviors and perceptions, you?ll want to report the average scores for all respondents, and then report the average for categories of people that responded differently. You?ll also want to report whether or not you think the perceptions are related in any way to the behaviors. In this section you should have some graphs embedded in the text which you?ll refer to in order to make sure that readers remember important facts (e.g., ?We found a difference between the responses of male and females. Figure 3 shows that females preferred the advertisement much more than males.?).
Recommendations (1 page maximum)
In this final section, you?ll give your recommendation. This is the answer to your research question. So, what do you think the company should do? Base your recommendation only on the data you?ve collected. This is a good time to review your top two or three findings as support of your recommendation.
Once you have recommended something, you also want to make sure you show the reader that you have thought through your recommendations. How do you expect the company to benefit based on your advice? Will they see it in terms of revenues, in more positive attitudes, or more frequent visits to the store? Some combination of these things? This is your last chance to convince the reader that it is worth following you recommendation.
Finally, you should inform the reader of the implications involved. If they accept your recommendation, what are they really signing-up for? For example, will your recommendation to move aggressively into the teenager segment mean that they will have to simultaneously lower prices and launch a communications campaign? What resources will have to be deployed to make your recommendation successful?

Appendix (about 2 pages, not included in page total)
This section will include the survey questionnaire and the raw scores for each question. You will list each question and the scale associated with it (i.e., 1-7 scale, agree-disagree; or yes-no, etc.). You have two options for reporting the raw results:
1. Give the average score for all your respondents (can only be done for 1-7 type scales)
2. Give the percentage of respondents that answered each category (should be done for category questions)
Use whatever formatting you think is easiest for the reader, but you?ll want it to include information like the following examples.

Q1. How likely are you to visit The Home Depot, next time you purchase gardening products? (1-7 scale, 1=Disagree strongly, 7= Agree strongly)
Mean=4.67, standard deviation=.92

Q4. To what extent do you agree with the following statements?
mean Standard deviation
A) Wal-Mart is ethical. 3.55 .87
B) Whenever Wal-Mart enters a city, local merchants suffer. 4.67 .92

Q9. Gender
Male: 43.6%
Female: 50.0%
Prefer not to answer: 6.4%

 
ORDER THIS ESSAY HERE NOW AND GET A DISCOUNT !!!

 

 

You can place an order similar to this with us. You are assured of an authentic custom paper delivered within the given deadline besides our 24/7 customer support all through.

 

[order_calculator]