Decision Support Systems

1
Decision Support Systems Marketing Research

• Joe Thomas
• (Joseph Thomas Paniculangara)

2
Decisions, Decisions, Decisions

• We make them all the time under uncertainty
• we do not know exactly what will happen as a
  result of our decision
• or we do not know what the results of our
  decision were
• In Marketing we conduct research to REDUCE not
  eliminate uncertainty

3
What did you do this weekend?

• Suppose you decided to go for a movie
• To make sure you got the most enjoyment from the
  movie you chose, you would have done some
  RESEARCH
• You used the Internet to get some information on
  the movies playing
• stars in the movie, genre type, length of movie,
  critics reviews, etc.
• SECONDARY research

4
Lots of uncertainty about the movie

• You werent sure, so you decided to get some
  information on your own
• PRIMARY research
• Next you asked your roommates over dinner what
  they thought of the movies playing
• FOCUS Group
• an example of qualitative research

5
But you dont like your roommates

• What type of people do you get information from?
• SAMPLING Sample is the small set of people you
  select from a population
• At least your focus group and secondary research
  allowed you to narrow your choice to
• Shrek or Pirates

6
Ask the people you do like

• You contact your closest friends by e-mail and
  ask them to rate the two movies
• SURVEY research conducted over the Internet
• You might also look at the ratings by viewers of
  the movies posted on websites
• Note that all the information you get is
  quantitative - numbers

7
A subject that you didnt like

• Statistics
• Levels of measurement
• Nominal, ordinal, interval, ratio
• Measures of central tendency (average)
• Mode, median, mean
• Measures of dispersion
• Standard deviation, variance, range
• Descriptive statistics
• Percentages, proportions, frequencies
• Inferential statistics
• t-tests, z-tests, etc. with their distributions

8
Do the results from your sample apply to you?

• You decide to watch the trailers of the two
  movies on the Internet and ask your closest
  friend to do the same
• Based on your and your friends reactions to the
  trailers, you try to predict which movie you will
  enjoy
• EXPERIMENTAL method something is manipulated by
  the researcher among people participating in the
  research

9
How do you make sense of all this information?

• You analyze the data using statistical methods in
  the case of quantitative data
• Different types of analysis for qualitative data
• You are now able to look at the results of the
  data and choose which movie to see
• The whole process has been expensive (time and
  money) but if this is the only movie you will see
  this summer

10
A few bonuses

• You remember your roommate said she/ he always
  took a potential partner to a romantic movie on
  the first date
• COMPETITIVE INTELLIGENCE
• Over time, you can build a computer-based system
  to help you predict which movies you will enjoy
• DECISION SUPPORT SYSTEM

11
Decision Support Systems

• An interactive, flexible computerized information
  system that enables managers to obtain and
  manipulate information as they are making
  decisions.
• interactive, flexible, discovery-oriented,
  accessible

12
Database Marketing

• The creation of a large computerized file of
  customers and potential customers profiles and
  purchase patterns
• The key tool for successful one-to-one marketing

13
Market Research

• Market research provides information about
  consumers
• The systematic gathering, recording and analysis
  of data with respect to a particular market,
  where market refers to a specific customer group
  in a specific geographic area
• Products, advertising, prices, packages, names,
  logos, services, buying habits, colors, uses,
  awareness, familiarity, new concepts, wants,
  needs, traffic patterns, politics, etc.

14
Market Research functions

• Descriptive function What is? research done
  on stable states
• Diagnostic function Why did? research done
  to identify how states changed as a result of
  things changing
• Predictive function What if? research done
  to predict how states will change as a result of
  changing things?

15
Uses of Market Research

• Improve the quality of decision making
• Trace problems
• Focus on keeping existing customers
• Understand the marketplace
• Alert them to marketplace trends
• Gauge the value of goods and services, and the
  level of customer satisfaction

16
Market Research Process
17
Key definitions

• Marketing Research Problem
• Determining what information is needed and how
  that information can be obtained efficiently and
  effectively
• Marketing Research Objective
• The specific information needed to solve a
  marketing research problem the objective should
  provide insightful decision-making information
• Management Decision Problem
• A broad-based problem that requires marketing
  research in order for managers to take proper
  actions

18
Secondary Research

• Secondary data is data previously collected for
  any purpose other than the one at hand
• Internal databases
• Government agencies
• Trade industry publications
• Business periodicals
• News media

19
Advantages

• Saves time and money if on target
• Aids in determining direction for primary data
  collection
• Pinpoints the kinds of people to approach
• Serves as a basis of comparison for other data

20
Disadvantages

• May not give adequate detailed information
• May not be on target with the research problem
• Quality and accuracy of data may pose a problem

21
Research Design
22
Primary Research

• Information collected for the first time. Can be
  used for solving the particular problem under
  investigation.
• Experimental, survey, observational qualitative
  methods

23
Advantages

• Answers a specific research question
• Data are current
• Source of data is known
• Secrecy can be maintained

24
Disadvantages

• Expensive
• Piggybacking may confuse respondents
• Quality declines if interviews are lengthy
• Reluctance to participate in lengthy interviews

25
Survey Methods

• Descriptive (not causal) what there is
• Use of a structured questionnaire to get
  information from respondents
• Sample of respondents drawn from a population
• Open-ended (respondents own words) vs.
  close-ended (respondent chooses between
  alternatives specified by researcher), scales
  also used

26
Types of Surveys

• Mall intercept
• Survey research method that involves interviewing
  people in the common areas of shopping malls
• In-home interviews
• Telephone interviews
• Mail surveys
• Internet surveys

27
Experiments

• Is there a significant causal effect of one
  variable on another
• Correlation effect must occur when the cause
  does
• effect must not occur when cause does not
• Sequence effect must follow cause and not the
  reverse
• Uniqueness effect must only occur when cause
  occurs

28
Experiments continued

• Independent variable must be amenable to
  measurement and manipulation
• Dependent variable must be able to be inferred
• To rule out confounds (other causes of the
  effect), a treatment (experimental) group is
  contrasted with a control group

29
Qualitative Methods

• Depth interviews
• Ethnographic methods
• Projective techniques
• Focus groups

30
Focus Groups

• Most popular mode of qualitative research
• Illusion of a large sample (8-12 participants)
  compared to interviews
• Participants sit around a table with the
  moderator at the head
• Camera and microphone
• One way glass behind which client sits
• Transcript of session
• About an hour long

31
Observational Methods
32
Scanner Data
33
Sampling

• A sampling frame is a list of people who are in
  the universe (population)
• A sample is a small set of people drawn from the
  universe
• The objective is to make inferences about the
  universe from the sample

34
Probability Samples

• A sample in which every element in the population
  has a known statistical likelihood of being
  selected
• Simple Random Sample, Stratified sample, Cluster
  sample, Systematic sample
• Sometimes a non-probability sample is used
• Convenience sample, Judgment sample, Quota
  sample, Snowball sample

35
Sampling errors
36
Data Analysis

• Depends on the data
• Content and text analysis for qualitative data
• Cross tabulations are the most popular analysis
  for survey data
• Banners on top, stubs on the sides
• Frequency counts are also popular

37
Research Presentation

• Concise statement of the research objectives
• Explanation of research design
• Summary of major findings
• Conclusion with recommendations

38
Internet in Marketing Research

• Allows better and faster decision making
• Improves ability to respond quickly to customer
  needs and market shifts
• Makes follow-up studies and tracking research
  easier
• Slashes labor- and time-intensive research
  activities and costs

39
Advantages of Internet Surveys
40
Samples available on the Internet
41
Online Focus Groups

• Build a database of respondents via Web site
  screening questionnaire
• Identify qualified individuals via e-mail
• Develop a discussion guide
• Moderator runs group by typing in questions
  online for all to see
• Environment is similar to a chat room
• Firm captures the complete text of the focus
  group

42
Steps in Market Research

• Necessity of on-going market research
• Panel
• Tracking
• Decision making in the next round
• Presenting the report
• Audience can follow logic
• Audience can follow technical parts

43
Competitive Intelligence

• An intelligence system that helps managers assess
  their competition and vendors in order to become
  more efficient and effective competitors.

44
Sources of Competitive Intelligence