Marketing
2. Consumer marketing research
Marketing
What are we gonna talk about today?
• Data - Information - Knowledge
• Marketing research process
• Baseline and sample
• Research typology
• Data collection methods and techniques
Data and information differences
• Data is the most basic form of knowledge, e.g. the
brand of butter sold to a particular customer in a
certain town. This statistic is of little worth in itself
but may become meaningful when combined with
other data.
• Information is a combination of data that provide
decision-relevant knowledge, e.g. the brand
preferences of customers in a certain age category
in a particular geographic region.
Goals of marketing research
•Understand the market
• Present
• Future
•Identify threads and opportunities
•Formulate marketing actions and assess its
effectiveness
Can I make a decision based on this data?
Age 2015 2016 2017
20-29 14% 12% 9%
30-39 17% 16% 13%
40-49 27% 23% 17%
50-59 20% 25% 22%
60-69 14% 15% 27%
70+ 8% 9% 12%
0%
5%
10%
15%
20%
25%
30%
20-29 30-39 40-49 50-59 60-69 70+
Age of customers
2015 2016 2017
7%
12%
17%
22%
27%
32%
20-29 30-39 40-49 50-59 60-69 70+
Age of customers
2015 2016 2017
Mohu udělat rozhodnutí na základě těchto dat?Is this data?
Case study: Campbell‘s
• Iconic American brand of canned food.
• The 143-year-old company has hit a major problem:
• A 13% drop in overall canned food consumption.
• A drop in market share from 67% to 53% due to the growing
popularity of fresh and premium soups.
• Campbell's identified a need to get to know the 18 to 34 year old
customer better, roughly 25% of the U.S. market and the future of the
company.
• Anthropological research was conducted by having company
representatives meet millennials (1986-1995) in their environment.
Case study: Campbell‘s
• Key information?
• Millennials love spicy food
• They eat more exotic foods than previous generations
• They don't have much time at home to cook
• Reactions?
• New product line of six exotic flavors
• Triple the price of a can
• Promoted online only
• Music and entertainment sites
• Gaming platforms
• Social media
• The result?
• Slowing the decline from 13% to 2%
Marketing research process
• PREPARATION PHASE
• Problem definition
• Situation analysis
• Research project planing
• Pre-research
• REALISATION PHASE
• Data collection
• Data processing
• Analysis and interpretation
• Visualisation and presentation
PROBLEM DEFINITION
• The most important part of
the research
• A well-defined problem is
half-solved
• There are no wrong answers
in marketing research, only
wrong questions
• Case study: a new range of
müsli bars
Case study: Flavours of müsli bars
• A cereal manufacturer wanted to enter the muesli bar market.
• He asked a research agency and commissioned research.
• He wanted to find out which of the three flavours would be
suitable for the market.
• The agency carried out taste tests with 200 customers.
• It recommended the banana flavour, which was the most
preferred by customers.
• The product was launched and failed completely.
• What went wrong?
Research question
• Research question is a clear statement of the problem
formulated in sentence ending with question mark.
• Clear expression of what will be studied (Kolb, 2008)
• It influence whole research process:
• What literature we have to study in advance?
• How we design research scope?
• What access to the date do we need to negotiate?
• What type of research design will be chosen?
• What data will be needed?
• Sample selection process.
• Nature of a final report.
Research question examples
•What?
•What make our clients leave for our competition?
•When?
•When do our clients visit our store for the first
rebuy?
•Where?
•Where exactly Instagram influencers buy fashion
products?
Research question examples
•Who?
•Who spreads negative brand sentiment online?
•How?
•How product rebranding influenced attitudes
of general public?
•Why?
•Why do companies outsource marketing
research?
Population and sample
• Would it be possible to survey all adults in the Czech Republic?
• Population: the set of all units covered by the research.
• Sample: a smaller part of the core population used to carry out
the research
Representativeness of the sample
•There are two basic types of selection
• Intentional - we decide in advance which units to
include in the research.
• Representative - we obtain the sample using accurate
mathematical models. This allows generalisation of
the results to a core set with clearly quantifiable
precision.
• Randomisation, tables, systematic sampling
Representativeness of the sample
•The first prerequisite is to ensure randomness of
selection.
•All members of the core group have an equal
chance of being selected.
•Case: Customers who visit the store 5 times more
often are 5 times more likely to be included in the
sample.
• Who visit store 5 times more often?
Sample and sample size
• Sample is a part of
population which, when
collected properly, can
produce results which are
generalizable.
• Greater the sample size
more reliable the results are.
• Reality = Parametr + Error
Population
Parametr
Data
Statistics
People
Sample
Favourite colour example
• In a class, teacher ask students what is their favourite colour.
• The realtime results goes like this:
Respondent Answer Results PINK in % Results GREEN in % Change
1 Pink 100 0 No data
2 Green 50 50 50%
3 Green 33,5 66,5 16,5%
4 Pink 50 50 16,5%
5 Green 40 60 10%
6 Green 33,5 66,5 6,5%
7 Green 28 72 5,5%
8 Pink 37 63 9%
9 Pink 44 56 7%
10 Pink 50 50 6%
Ideal sample size
!
"!
#!!
#"!
$!!
$"!
%!!
%"!
! "!! #!!! #"!! $!!! $"!! %!!! %"!! &!!! &"!! "!!!
'()*+,-(./+*)0(1
2*P.+4,50S4
Sample size in qualitative studies
• The data collection takes place as
long as there is no condition in
which further examination of the
selected sample does not bring
new substantial information.
• Theoretical saturation
• The goal of qualitative research is
not generalization, so it is not the
aim of the results to relate to the
whole population but to reveal the
connections and causes of a
certain behavior of customers.
Quota sampling
• This is a quasi-representative technique - quotas are set
subjectively - it depends on the qualities of the
researcher.
• We use our own thinking to select the traits by which we
meet quotas.
• This technique is intended to ensure similarity between
the sample and the base population.
Errors in data collection
•Sampling errors
•Certain variations at different sample sizes
must be taken into account
•Varies with sample size
•Non-sampling errors
•Misunderstood question, refusal of interviews,
missing observations, coding errors.
Marketing research:
In-house or agency?
BRIEF
RESEARCH PLAN
Business
problem
Know-how
ORDER
RESEARCH REPORT
$$$ $$$ $$$
Knowledge
Typology of marketing research
Quantitative / Qualitative
Primary data / Secondary data
Exploratory / Causal (Explanatory) /
Descriptive
Research types
• Quantitative
• Based on numbers
• Measurement and counting
• Large sample size
• Qualitative
• Based on text, sound, video and pictures
• Interpretation
• Small sample sizes
DATA
Secondary Primary
ObservationSurvey ExperimentExternal Internal
Studies of
associations,
chambers of
commerce, Czech
trade, CSU, trade
magazines and
media,...
Customer reports,
call centre reports,
financial
statements,
previous research,
databases,...
Individual
interview,
group
interviews,
online and
offline
questionnaires
Mechanical
and electronic
accesses,
personal
accesses
Controlled
experiment,
quasi
experiment,
field
experiment
Studijní opora strana 102 - 104
•Exploratory research: what is
there? What's actually going
on?
•Causal (explanatory)
research: How did it happen?
What caused it?
•Descriptive research: How
much is there? How big is it?
Studijní opora strana 102 - 104
•Exploratory research: What's
the next series we should
produce?
•Causal Research: Which
homepage design leads users
to view more content?
•Descriptive research: Which
series is the most popular?
PRIMARY DATA
COLLECTION METHODS
Observation
•Overt
•Persons are informed that research is being
conducted.
•There may be bias due to distortion of natural
behaviour.
•Hidden
•Subjects are unaware of the research.
Observation
•Participatory
•The observer joins and becomes part of the group.
•Participates in its life and activities.
•It is very important not to interfere too much in the
dynamics of the group.
•Non-participatory
•The observer is outside the group.
Experiment
• Finding a causal relationship between variables.
• Dependent and independent variable.
Survey
•Questionnaire
• Online/offline
• Self-
completion/completion
during interview
•Interview
• Structured/Non-
structured
• Individual/Grou
Case study: Pepsi challenge
• The famous Pepsi challenge campaign, which included a taste experiment,
showed that people preferred Pepsi to Coca Cola.
• What could have caused the failure of The New Coke, which was Coke's
response to the market situation?
• Independent variable: the variable that causes the change. In
an experiment, it is the variable we manipulate.
• Dependent variable: The variable that is affected by changes
in the independent variable.
Variable types
Independent
variable
Dependent
variable
• A correlation has been found between breast implants and
suicide (Koot, Peeters, Granath, Grobbee, & Nyren, 2003).
• However, we cannot claim that breast augmentation surgery
results in suicide.
• There is another external factor (another variable) that
causes both phenomena - that is self-esteem.
Causality problems
Causality problems
• First causal model: the more plastic surgery, the more suicides
• The second causal model: the more problems people have with their
self-esteem, the more plastic surgery and suicides.
Plastic surgery Suicide
Plastic surgery
Suicide
Self-esteem
problem
Quality measures
•Representativeness:
•Indicates the relationship between the sample and
the base population
•Research quality
•is determined by validity and reliability
Validity
Valid questions are those that provide
answers to exactly what we are asking - what
is the main objective of the research.
Reliability
It expresses the degree of stability of the
research tools.
The extent to which the question remains
reliable and valid in subsequent iterations for
example, in different temporal, social and
cultural settings.
New trends in marketing research
Thank you for your attention