Marketing Research

1
Marketing Research

• David Forlani
• University of Colorado at Denver and Health
  Sciences Center

2
Presentation OverviewMarketing Research

• What is it?
• Why is it conducted?
• How is it done?

3
Defined

• Marketing Research is the planning, collection
  and analysis of data relevant to marketing
  decision making and the communication of the
  findings to management.
•  OR
• Marketing Research is a tool for improving the
  quality (financial viability) of marketing
  decisions.

4
3 General Types

• 1. Programmatic ResearchOngoing intelligence
  gathering process, e.g., alerts product managers
  to cultural and attitudinal trends, new market
  opportunities, methods of better reaching
  markets.
•  2. Selective ResearchDone to test the
  viability of an idea, e.g., product concept,
  packaging design, ad copy.
•  3. Evaluative ResearchUsed to assess a prior
  decisions outcomes, e.g., how effective was the
  new ad campaign, is the new distributor in
  Arizona reaching more potential users?

5
3 Functional Applications

• 1. Descriptivegathering and presenting
  statements of fact, e.g., industry sales trends,
  competitors advertising levels or consumers'
  brand attitudes.
• 2. Diagnosticexplaining outcomes or events,
  e.g., assessing the profit impact of switching
  from a direct to a broker-based saleforce? or why
  did 3rd quarter sales fall in the northeast?
•  3. Predictivewhat will be the effect of an
  anticipated marketing action, e.g., how will a
  change in packaging design impact user
  satisfaction and brand equity?

6
Basic 10-Step Process

• 1. Problem Identification
• Why is marketing research being considered?
• Is this a problem that it can solve?
• What decisions can be made after the research is
  done that cannot be made/made well before hand?
• What is the cost/benefit ratio given the amount
  of time and money involved?

7
Basic 10-Step Process

• 2. Problem Specification
• Do we really understand the nature of the problem
  being investigated?
• Are we clear on what contribution marketing
  research can make?
• Is Primary research really necessary?

8
Basic 10-Step Process

• 3. Set Research Design
• What type of information is needed to address the
  Research Question?
• If Qualitative, are we interested in Concepts,
  Attributes, Motives, etc.?
• If Quantitative, are we interested in
  understanding existing cognitions like Beliefs,
  Attitudes, or Preferences, or in determining
  Causation?

9
Basic 10-Step Process

• 4. Choose the Method of Study
• What types of decisions are being made?
• Focus Groups can provide insights into what
  people see as important when evaluating concepts.
• Surveys can provide descriptive information about
  who people are, what they think, and what they
  think they do.
• Observation provides information about what
  people actually do.
• Experiments can demonstrate cause and effect
  linkages.

10
Questionnaire Design Format

• Determine Surveys Frame of Reference
• Last purchase, Next purchase, Most people, You, A
  Close Friend, A Child/Parent
• Create Question Stems
• Direct, Comparative, Situational
• Set Response Format
• Open/Closed Ended
• Categorical/Continuous
• Level of Measurement

11
Questionnaire Design Wording

• Clarity Where do you usually shop for clothes?
• Unbiased How often do you shop at low-class
  stores like K-Mart?
• Vernacular Quantify the degree of variability
  that currently exists in your propensity to view
  store choice as a malleable construct?
• Sensitivity How often do you shoplift when you
  feel that a store's prices are too high?

12
Questionnaire Design Flow

• Statement of Purpose I am working for
• Instructions Please circle the number
• Screening Questions In the last month, have you
  eaten
• Focal Questions Those that answer the Research
  Question(s).
• Tracking Questions Demographic and Psychographic
  identifiers of the respondents.
• Signal the End of the Survey THANK YOU!

13
Questionnaire Design New Product Essentials, How
Many How Much

• How likely is the respondent to buy your
  innovation?
• If a new beverage as described above were
  available how likely would you be to buy it?
• Extremely Likely 1 2 3 4 5 6 7 8
  Extremely Unlikely
• How much will s/he pay?
• Compared to other products like ? ? ?, for the
  product described above, would you be willing to
  pay
• Much More 1 2 3 4 5 Much Less
• How much would you be willing to pay for this
  product
• ___lt1 ___ 1-1.50 ___ 1.51-2 ___ 2-3
  ___gt3

14
Basic 10-Step Process

• 5. Design the Study
• Have we properly specified the constructs of
  interest? E.g., satisfaction, brand attitude.
• Do our measures fully uniquely capture these
  constructs? E.g., Do they cover constructs
  domain and just its domain?
• Are the measures reliableinternally consistent
  and temporally stable?
• Has the opportunity for systematic error been
  controlled? E.g., controlled for bias in question
  design and response formats.

15
Basic 10-Step Process

• 6. Develop a Sampling Plan
• Probability samples allow us to estimate sampling
  error. Thus we can project our results to a
  larger population with confidence.
• Non-probability samples do not, but may be fine
  for investigating relationships between variables.

16
Probability-Type Samples

• Simple Random use a random number generator
  giving all members of the population an equal
  chance of being selected.
• Systematic random start, then use a fixed skip
  interval.
• Stratified simple random within strata (levels
  of a variableage) on a proportional or
  disproportional basis.
• Cluster sampling unit is a group instead of an
  individual (counties versus citizens).

17
Non-Probability Type Samples

• Convenience choose the most available subjects.
• Judgment identify appropriate characteristics to
  sample for.
• Quota ensure representation of underrepresented
  subjects with key characteristics.
• Snowball sampling by referral. When the
  incidence of usage in the general population is
  tiny.

18
Basic 10-Step Process

• 7. Collection of the Data
• Establish clear procedures for data collection,
  ones that minimize the introduction of procedural
  bias.
• Are we speaking to the target respondent?
• Did we approach people as indicated in the
  sampling plan?
• Are we rotating response choices to eliminate
  order-effects.

19
Basic 10-Step Process

• 8. Analysis of the Data
• Have the proper statistical techniques been
  applied?
• When are Linear Regression, Logistic Regression
  and Discriminant Analysis appropriate?
• Are we over-interpreting the data?
• Do the results address the original research
  questions?

20
Basic 10-Step Process

• 9. Writing and Presenting the Results
• Have we properly communicated the statistical
  results?
• Do all relevant parties understand the
  implications of the results?
• With how much more confidence can the decisions
  that motivated the research be made?

21
Basic 10-Step Process

• 10. Follow-up
• How confident are we in the accuracy of the
  research findings?
• What decisions have been made based on the
  research effort?
• What other factors might be considered in future
  research efforts?
• If the action investigated is pursued, has a
  follow-up study been scheduled to assess its
  outcomes?

22
In-Class ExerciseFor Your Group Project

• Identify questions to which you need answers
  before moving forward.
• Sort them by source of data
• Secondary and Primary.
• Outline an approach for obtaining the primary
  data.
• State Objectives and Sketch out a Survey.
• Key items How much and How many?