Marketing Information Systems

1
Marketing Information Systems

• Better Marketing Decisions Through Research

Creativity is more important than knowledge A.
Einstein
2
Marketing Information Systems
3
Marketing Information Systems

• INFORMATION
• Customers
• Competitors
• Vendors
• Environment

4
Types and Sources of Data

• Types
• Primary
• Secondary
• Sources
• Internal
• External

5
Sources Types of Data
Source Internal External
Type Secondary Primary
Sales Records
6
External Secondary Data

• Governmental Sources
• U.S. Census of Population
• U.S. Census of Retailing
• U.S. Industrial Outlook

7
External Secondary Data

• Private Data
• Consumer Panels
• Panel Data Data collected over time from same
  households
• Scanner Panel Data Data collected using retail
  scanner technology
• Suppliers
• A.C. Nielsens SCANTRAK (gt40 k HH)
• Information Resources Inc. (gt60 k HH)

8
External Secondary Data

• Private Data (Contd)
• Media and Audience Data
• Simmons readership surveys
• Leading National Advertisers (LNA)
• Neilson National Television Index
• Sales Data
• Nielsen Retail Index

9
Advantages and Disadvantages of Secondary Data
ADVANTAGES
DISADVANTAGES
DATA
Data may be obsolete
Government Data
Data may be inappropriately classified
Industry Data
Data may be incorrect
10
Stages in the Research Process

• 1. Define the Problem/Opportunity
• 2. Conduct Situational Analysis
• 3. Conduct Exploratory Research
• 4. Formulate Research Questions Predictions
• 5. Design Study
• 6. Collect Data
• 7. Analyze and Interpret Data
• 8. Present Findings and Recommendations

11
Stages in the Research Process

• 1. Define the Problem/Opportunity
• Symptoms?
• Potential Causes?
• Core Research Objectives and Questions?

12
Stages in the Research Process

• 2. Situational Analysis
• Existing Information
• 3. Exploratory Research
• Key Informant Interviews
• Focus Groups
• Depth Interviews

13
Stages in the Research Process

• 4. Formulate Research Questions Predictions
• Predictions are called Hypotheses
• May have a set of detailed research questions
• Evolved from the objectives of Step 1 and the
  information and insights of Steps 2 and 3

14
Stages in the Research Process

• 5. Design Study (Primary Data Collection)
• Define the population of interest
• Determine the data collection method
• Determine the sampling method

15
Stages in the Research Process

• 6. Data Collection
• Observation
• Focus Group
• Surveys
• Experiments

16
Observation

• Human
• Phantom Shoppers (KFC)
• Mall Enumerators

17
Observation

• Human
• Phantom Shoppers (KFC)
• Mall Enumerators
• Mechanical
• Retail scanners
• People meters
• Observation generally tells use what but not why
• Exception Single source data systems

18
_____________ Data Systems
Household
Retail Scanner
Television Monitor
19
Focus Group Interviews

• A popular type of QUALITATIVE research
• Small group (6 to 10 people) - 1 hour
• Group leader guides discussion
• Designed for in-depth, open-ended responses
• (-)
• ()

20
Survey Method

• A questions-based method
• Can be qualitative or quantitative
• Collect information from a SAMPLE and draw
  conclusions about the POPULATION
• Attitudes
• Beliefs
• Product-related Behaviors
• Mail/Phone/Personal/Computer-based

21
Sample Attitude Questions

• Quantitative On the scale below, indicate how
  much you like IBM computers
• Qualitative Please tell us about your feelings
  toward IBM computers

1 2 3 4 5 6 7
Dislike
Like
22
Methods of Conducting Interviews
Personal Interview
Web Research
Mail
Telephone
23
Ratings of Different Data Collection Methods
Factor Telephone Mail
Personal

• 1. Information Quantity Limited
  Limited Excellent
• 2. Control of Sample Excellent Poor
  Good
• 3. Response Rates Higher
  Lower Higher
• 4. Supervision Excellent
  None Fair
• 5. Speed of Collection Faster
  Slower Slower
• 6. Cost Lower
  Lower Higher
• 7. Flexibility Medium
  Low High

24
Information Quantity
Telephone Mail Personal Limited
Limited Excellent

• Long telephone/mail causes lower response
• Poorly designed telephone/mail causes lower
  response

25
Control of Sample
Telephone Mail Personal Excellent
Poor Good

• Telephone Allows call backs and screening.
• Mail Nonresponses and self selection can cause
  sample bias

26
Response Rates
Telephone Mail Personal Higher
Lower Higher

• Social pressures to participate
• Recent concerns about telephone
• Unlisted Numbers Call Screening
• Do Not Call Registry???
• How to increase response rates for mail
• Length
• Incentives Sponsors
• Reminder Cards Return Envelopes

27
Ratings of Web-based Surveys
Factor Web-based Surveys

1. Information Quantity Limited (like mail
and telephone) 2. Control of Sample ????
(internet vs. noninternet diff) 3.
Response Rates Lowest (in general)
4. Supervision None (like
mail surveys) 5. Speed of
Collection 6. Cost
7. Flexibility
http//www.zoomerang.com (web-survey
site)
28
Experiments

• MANIPULATE one variable at different levels to
  find its effect
• Advertising
• Price
• HOLD CONSTANT everything else
• MEASURE effects using observation or questions

29
Lab vs. Field Experiments

• Field experiments conducted in more realistic
  settings but give up control
• New Product Experiment
• Lab Experiment Colgate tested Soaky under all
  water conditions
• Field Study Real kids and the relaxing
  effects of warm water.
• Test Marketing

30
Sampling Issues

• Why Sample?
• Defining the Target Population
• Representativeness
• Sampling Methods

31
Sampling

• Why Sample?
• Time
• Money
• Accuracy
• Example A.C. Nielsen
• Samples 3000 retail outlets in U.S.
• Estimates Cokes NATIONAL market share

32
Sampling Issues

• Defining the Target Population
• What are their characteristics?
• Demographic
• Psychographic
• Product-related

33
Sampling Issues

• Representativeness

• Does sample accurately reflect population?
• Affects the validity of your conclusions
• Examples
• 1. Nielsens cola data
• Small retailers only??
• Restaurant sales only??

34
Sampling Issues

• Representativeness

• Examples
• 2. Customer satisfaction mail survey
• Respondents may represent extreme attitudes
• Nonrespondents may be reasonably satisfied

35
Sampling Issues

• Sampling Methods

• a. Nonprobability
• Convenience
• b. Probability
• Simple Random
• Stratified Random

36
Convenience Sample

• Ask the first and most convenient people you can
  find
• Easy...but are they representative of the
  population of interest?
• Survey of student concerns from convenience
  sample of 100 students eating at Ferguson Center

37
Simple Random Sample

• Each member of population has EQUAL, KNOWN, and
  NONZERO chance of being in sample
• EQUAL
• KNOWN
• NONZERO
• Good when the population is fairly homogenous to
  begin with

38
Stratified Random Sample

• Used when distinct groups (strata) within a
  population need to be represented
• Random sample might under-represent group by
  chance
• Stratified random ensures representation of each
  group

39
Stratified Random Sample

• Process
• SORT by the grouping variable
• SELECT a random sample within each group
• Example Pharmaceutical Company
• How often do physicians prescribe Drug X?
• M.D.s (70)
• D.O.s (30)
• Sample proportionately to ensure representative
  sample