Is time availabile before a decision must be made?

Is information already on hand inadequate to make a decision?

Is the decision of strategic or tactical importance?

Does the value of the research exceed the costs?

If no to any of these, don't conduct research!

Benefits: decreased uncertainty, increased likelihood of correct decision, improved marketing performance/higher profits

Costs: Research expenditures, delay of decision and disclosure of information to rivals, possible erroneous research results

1. formulate problem

2. Determine research design

3. Design data collection method and forms

4. design sample and collect data

5. analyze and interpret data

6. prepare the research report

Big bucks

Big decision

Survey myopia

Sophisticated Researcher

Basic research is to increase knowledge generally

Applied research is to solve a specific problem

Exploratory - ambiguous problem

Descriptive - Aware of problem

Causal - Problem clearly defined

1. decision problem

2. research problem (objective)

3. research question

4. possible variables to measure

discovery of ideas/insights

ex: what are the most important drivers of customer satisfaction?

deals with frequencies and relationships between variables and typically guided by hypotheses

ex: How can we redesign customer satisfaction programs to improve our market share

concerned with cause and effect relationships, most often through experiments

ex: which satisfaction programs will be most successful in our target markets?

1. Ascertain the decision maker's objectives

2. understand background of problem

3. isolate/identify problem, not symptoms

4. determine unit of analysis

5. determine relavent variables

6. state research questions and objectives

Primary = any original research done by you

Secondary = any data collected from anywhere else/a previous study

Internal sources: sales results, customer feedback, price info, scanner info

External sources: published data, internet

Advantages/Disadvantages of secondary data

Advantages: quick to obtain, inexpensive (?), information not otherwise accessible, might be unnecessary, help redefine research problem

Disadvantages: lack of availability, lack of fit, lack of accuracy, insufficient data

Quantitative: Asking questions with specific answers to many participants

Qualitative: Asking more open-ended questions to fewer participants

Pros: group dynamics (synergy), security/comfort for participants, quick/cheap, rich/flexible data, useful for non-literate groups (kids)

Cons: can't generalize, "focus groupies", group dynamics, biased moderator, difficult to summarize, easy to "cherry pick"

unknown population (new insights)

smaller sample/not random participants

results might not generalize to other people

Lack focus group interaction (but also the pressure of one)

word association

sentence completion

cartoon test

third person/role playing

consumer drawings

thematic apperception test (picture interpretation)

information is observable/inferable

behavior of interest is repetitive/predictable

behavior of interest is of short duration

human behavior/physical action

verbal behavior

spatial relations/locations

physical objects

temporal patterns

verbal/pictoral records

expressive behavior

natural vs. contrived

open vs. disguised

structured vs. unstructured

direct vs. indirect

human vs. mechanical

Observe: see what you see, may influence behavior, not all things can be observed, can be quicker

Ask: rely on consumer memory, good for "why" questions, costly, responses might be inaccurate

systematic errors: respondent biases (non-response/self-affirming) and administrative biases (data processing, interviewing, sample selection)

Random sampling error (unavoidable)

Person administered (door-to-door or mall)

phone administered (traditional or computerized)

self administered (mail or web)

Advantages: can get virtually any type of info, versatility, broad coverage of respondents

Disadvantages: limited info from each person, not as objective as observation, must plan to avoid errors

descriptive studies:

Longitudinal (true panel, omnibus panel) and cross-sectional (sample survey)

concomitant variation (correlation): evidence of association between x and y

temporal antecedence: x must occur before y

no third factor driving both x and y

artificial correlation

danger of data mining

lots fo things are correlated without cause

1. one or more independent variables are manipulated

2. data on the outcome is gathered

3. other variables that may influence the effect are controlled

internal: did the treatment cause the effect?/am i measuring what I think I'm measuring?

external: can you draw conclusions about a larger conclusion?

Between - each only receives one treatment, comparisons made across groups

Within - subject gets more than one treatment, comparisons made on same subject

o: any formal observation/measurement

x: exposure of experimental units to treatment

eg: experimental group

cg: control group

r: random assigment