Académique Documents
Professionnel Documents
Culture Documents
Report by YHL
06/23/2010
normal
Reference Book
normal
normal
normal
3. Descriptive Research
4. Experimental Research
normal
Introduction
Why Marketing Research?
Marketing research is pervasive. Its the foundation for many business decisions.
normal
3 Types of Research
1. Exploratory Research
Secondary Research Observational Research Focus Groups Survey Research Factor and Cluster Analysis Preference Scaling Conjoint Analysis Marketing Mix Testing Concept Testing and STMs Field Experiments
7
2. Descriptive Research
3. Experimental Research
normal
normal
3. Descriptive Research
4. Experimental Research
normal
Overview
Secondary Research
Definition Advantages and disadvantages Sources of secondary data
10
normal
Secondary Research
Definition
Statistics not gathered for the immediate research problem but for some other purpose
Advantages
Secondary research is fast and cheap Easiest way to cut research costs
Disadvantages
Questionable reliability and validity Over aggregated Out of date More complete information on own company than competitors
11
normal
Library Sources
Books and periodicals Computerized databases
Non-library Sources
Trade associations and media Government agencies Syndicated services Personal networking Internet and WWW
12
normal
13
normal
Customer Information
14
normal
Environment
Economic indicators Political and social trends Technology trends
15
normal
3. Descriptive Research
4. Experimental Research
normal
Examples
Did you wash your hands after using the bathroom? How much television did you watch last week? How do you shampoo your hair?
17
normal
Observational Research
Research Techniques
Shadowing the consumer during the shopping or consumption experience Observation followed by interview Ethnographic immersion Train respondent to be amateur ethnographer
Observer types
Human beings Computer
18
normal
19
normal
20
normal
21
normal
Speed
Real time data (e.g., for staffing, replenishment)
Data Integration
Link path, penetration, conversion data to consumer demographics, shopping basket, purchase history
22
normal
23
normal
Traffic Flow
Identify dominant pathways through the store Angle and direction of approach determines best position/orientation for signs and displays. The greater the speed of approach, the shorter the message Facilitate incoming access to destination products, outgoing access to impulse items
normal
2.
Measure category penetration, dwell time, and conversion Measure line queues and crowding Cluster shoppers based on path similarity
3.
4.
Behavioral segmentation
25
normal
3. Descriptive Research
4. Experimental Research
normal
Focus Groups
Characteristics
Exploratory, qualitative research Six to twelve participants per group Similar demographics within a group Current customers and/or future prospects Informal discussion - 1 to 2 hours Objective moderator leads discussion Outline of topics from general to specific
27
normal
28
normal
Disadvantages
29
normal
30
normal
normal
32
normal
3. Descriptive Research
4. Experimental Research
normal
Survey Research
Questionnaire Development Process
Determine survey objectives, resources Determine data collection method Determine question response format Determine question wording
Revise as needed
34
normal
35
normal
Mail Questionnaire Cost Response Rate Sample Quality Questionnaire Length Visual Aids Speed Anonymity Requires Reading & Writing Low Low Low Depends on motivation Some Low Low Yes
Online Survey Low Low (High for Panel) Low (High for Panel) Short Some High High (Low for Panel) Yes
36
normal
Simple to administer and code. May oversimplify response options. Can allow for a broad range of possible May alert respondents to unfamiliar responses. response options. Allows for degree of intensity and feelings to be expressed. Simple to administer and code. Respondents may not relate to scale. Scale may be forced or overly detailed.
37
Scaled response
normal
38
normal
Complementary Alternatives
Would you say its better to regulate business pretty closely, or would you say the less regulation of business the better? OR Would you say its better to regulate business pretty closely or fairly loosely?
39
normal
Balance
Which one of the following would you say is most important to you in buying a new hat? [ ] style [ ] good looks [ ] workmanship [ ] material [ ] appearance
40
normal
How Specific?
What percent profit would you say that your company made last year? OR About how much after tax profit as a percent of sales would you say that your firm made last year?
Extensive Questions
How many eggs do you fry in a year? OR About how many eggs do you fry in a typical week?
41
normal
42
normal
43
normal
44
normal
Sample size
What size sample do I need? Influenced by a number of factors, including the purpose of the study, population size, the risk of selecting a "bad" sample, and the allowable sampling error.
Confidence interval (the level of precision) Confidence level (risk level)
normal
Sample Representative
Test whether samples can represent population in terms of different composition, such as gender, education, income, area, age, etc
Chi-square test: c2 = [(Oij - Eij)2 / Eij] where: Oij = the observed frequency for the cell in row i and column j Eij = the expected frequency for the cell in row i and column j degrees of freedom (df) = (rows - 1) * (columns - 1) probability of relationship by chance = chidist(c2, df) In SPSS: use nonparametric tests chi-square In Excel: use formula, CHITEST
46
normal
Common Statistics
Nominal and ordinal
Frequency counts, percentage,in terms of age, gender, education, area, profession,
Statistics
Mean, standard deviation, standard error, variance
Regression analysis:
Y=0+1*X1
47
normal
3. Descriptive Research
4. Experimental Research
normal
Market Segmentation
Market segmentation is concerned with identifying and capitalizing on individual or inter-group differences in product demand and response to marketing mix variables Through research, we seek to identify distinct segments with homogeneous needs, values and usage profiles Segmentation variables
Demographics/geographics Attitudes, opinions and interests Psychographics/lifestyles Benefits desired Brand usage and loyalty
49
normal
Factor Analysis
Reduces a large set of correlated (non-orthogonal) variables to a smaller set of uncorrelated (orthogonal) factors Estimates factor scores for each respondent representing the individuals performance on the underlying factors Application
Derive dimensions of product or service evaluation from a set of attribute ratings Derive customer lifestyle dimensions from a set of attitude, opinion, and interest statements Eliminate multicollinearity before conducting regression or cluster analysis
50
normal
51
normal
Apply cluster analysis to one or more quantitative measures to group respondents into a specified number of clusters Try reducing or increasing the number of clusters to equalize the size of each cluster and the distance between cluster centroids Select the number of clusters which is most easily interpreted and managerially useful Interpret the differences between clusters by examining the cluster means (centroids) for each of the input variables. Crosstabulate cluster membership with customer descriptors, attitude ratings, and purchase measures Concerns
Cluster analysis assumes that the input variables are uncorrelated (orthogonal) If the input variables use different scales of measurement, the values should be standardized before cluster analysis. Selecting the number and names of clusters is a subjective process. These decisions will affect the outcome of the analysis.
52
normal
3. Descriptive Research
4. Experimental Research
normal
54
normal
Respondents can
Rate the degree of similarity between all possible pairs of products Rate random subsets of the total set of pairs Place sets of similar products into groups
55
normal
56
normal
3. Descriptive Research
4. Experimental Research
normal
Overview
What is conjoint analysis? Why is it so popular? Applications of conjoint analysis An example Designing a conjoint study Looking ahead
58
normal
Conjoint Analysis
A method for understanding how consumers make trade-offs among features and/or characteristics of a product or service A tool for exploring the implications of such trade-offs for the design of new product and service offerings Assumptions
Each product is made up of a bundle of attributes; each attribute having two or more levels A consumers overall evaluation of a product or service is the summation of the utilities (or part-worths) of each of the product or services attributes
59
normal
Outputs
Individual-level estimates of the importance of each product or service attribute Individual-level estimates of the utilities (part-worths) of each level of each attribute Market share estimates
60
normal
61
normal
Conjoint Applications
Consumer Goods and Services
Bar soaps, hotels, auto tires, hair shampoos, employee benefit packages, carpet cleaners, gasoline pricing, panty hose, MBA job offers, credit card features, car rental agencies
62
normal
1.
Identify actionable attributes through expert judgment, focus groups, or consumer surveys. Use attributes that consumers can relate to. Attributes can include physical characteristics, features or options, product or package appearance, brand name, price, warranty, advertising, promotion, distribution, and bundled services.
63
normal
64
normal
65
normal
66
normal
Select a design by
Clicking on *Select Design+ Select a design from the list shown at the bottom of the screen (e.g., 1/2 fraction) View the design by clicking on *Design Details+ and then select the Design Listing tab from the top of the screen Each row of the design represents a hypothetical product that consumers will rate
67
normal
Ask consumers to evaluate profiles consisting of combinations of attribute levels. Each profile represents a product or service concept. Consumers either rank-order the profiles, rate profiles (e.g., on a scale of 1 to 100), or choose between pairs of profiles (paired comparison).
68
normal
69
normal
Run the SAS Market Research Application* Compute the relative importance of each attribute Compute the utilities associated with each level of each attribute Perform simulations to identify the best product designs
*See the reading Introducing the Market Research Analysis Application for additional details.
70
normal
The product or service category is familiar Stimuli realistically portray the concepts Attribute combinations are believable The attributes are relatively independent The true decision rule is compensatory Consumers evaluate a limited number of concepts
71
normal
Looking Ahead
Experimental Research
Simulated Test Marketing Field Experiments
72
normal
Descriptive Research
Survey Research Factor and Cluster Analysis Preference Scaling
Experimental Research
Conjoint Analysis
73
normal
Overview
Evaluating a new product concept Composition of a concept description Identifying the drivers of preference Looking Ahead
74
normal
How much do customers like the concept? Who likes (or dislikes) it? Why do they like (or dislike) it?
75
normal
Present various concept descriptions to customers in selected segments Evaluate concepts on purchase intent:
[ [ [ [ [ ] ] ] ] ] Definitely would buy Probably would buy May or may not buy Probably would not buy Definitely would not buy
Top two box
76
normal
Structure
Product attributes: color, texture shape, smell, taste, etc.
Function
What it does; how its used
Psychological
Fit with consumer lifestyle; personal benefits
Social
Interpersonal benefits
Economic
How much does it cost? How much money will it
77
normal
Pastato Romano
Pastato Romano is a delicious new pasta dish based upon the classic Italian pasta called gnocchi. Gnocchi is a traditional flour pasta that is filled with potato to make a light, yet satisfying side dish. Many people think of pasta as a main dish or appetizer; but perfect, piping hot pasta is also one of the best ways to make your meal complete. Past is very versatile and complements chicken, beef, pork, or fish. From Milan to Rome, pastas take on the unique flavors of the various regions of Italy. One of our
78
normal
Compute the percent of consumers who checked the top one or two boxes. Pick concepts that have the highest ratings. Alternatively, compare ratings to norms established in each product category. In addition to purchase intent, it is common to measure concept uniqueness and value.
79
normal
Pastato Romano
How likely would you be to purchase this product?
Definitely would buy Probably would buy May or may not buy Probably would not buy Definitely would not buy 10% 40% 27% 13% 10%
80
normal
81
normal
Attribute ratings:
Ask consumers to rate importance of attributes, or use regression to estimate importance from overall attitudes (Y) and attribute ratings (Xs).
Conjoint analysis
Derive attribute utilities (or part-worths) by asking 82
normal
Likes: Dislikes: Like gnocchi Prefer to make this myself Love gnocchi Prefer to prepare Seems tasty own sauces Love pasta; simple; quick Prefer fresh Love Italian tomatoes ingredients Like basil and tomatoes Like fresh tomatoes and basil Will not taste like Sounds delicious; wouldnt buy if it had pasta homemade chemicals
83