Lecture 1: Foundations & Data Classification
MUST KNOW
What Is Marketing Research? The systematic planning, collection,
analysis, and communication of data relevant to marketing decision
making.
Purpose:
To reduce uncertainty in managerial decisions.
To improve performance and profitability by supporting data-
driven insights.
Marketing Decision Problem vs. Marketing Research Problem
Marketing Decision Marketing Research Problem
Problem
Action-oriented (“What Information-oriented (“What do we need
should we do?”) to know?”)
Focuses on symptoms Focuses on underlying causes
Example: “Why are sales Example: “What factors influence
declining?” customer satisfaction?”
This distinction is essential for defining good research questions.
Einstein’s quote fits perfectly here: “The formulation of the problem is
often more essential than its solution.”
The Iceberg Principle
“What you see (the symptoms) is only the tip of the iceberg; the true
causes lie beneath the surface.”
Managers often react to visible issues (declining sales, poor
campaign performance) without identifying underlying causes
(brand perception, pricing, targeting).
Good marketing analytics digs deeper to uncover why.
,4. Types of Data (Quantitative vs. Qualitative)
Qualitative Data Quantitative Data
Non-numerical insights (text, Numerical data (can be measured
opinions, emotions) statistically)
Useful for exploring motives, Useful for measuring patterns,
emotions, ideas testing hypotheses
Examples: Focus groups, Examples: Surveys, experiments,
interviews, text analysis customer databases
Hybrid skills are increasingly important: combine intuition (qual) with
statistical evidence (quant).
Levels of Measurement (super important for data analysis)
Level Definition Example Statistical
Techniques
Nomin Categories with no order Gender Chi-square test
al (male/female)
Ordina Ordered categories, but Clothing size (S, Median, rank-
l no consistent spacing M, L) order tests
Interv Ordered, equal spacing, “Need for Mean,
al no true zero uniqueness” scale correlation
Ratio Equal intervals + true Company profit, Regression, t-
zero sales test
Why it matters:
The level of measurement determines which statistical tests you can apply.
Types of Marketing Research (by Research Design)
Type Goal Example Common
Techniques
Exploratory Gain insight, Focus groups, Qualitative
clarify ideas interviews, literature methods
review
Descriptive Describe “Do larger stores Surveys,
& relationships, test generate more regression
Predictive hypotheses sales?” analysis
Causal Test cause–effect “Does store size Experiments,
relationships cause higher sales?” ANOVA, t-test
(mediation and
moderation)
,Primary vs. Secondary Data
Primary Data Secondary Data
Collected for the current research Already collected for another
purpose
Examples: Surveys, interviews, Examples: Company reports,
experiments customer databases, industry data
More expensive but specific Cheaper but less tailored
Two types of secondary data:
Internal (sales records, CRM data)
and External (market reports,
syndicated data like Nielsen, GfK).
Syndicated Research
Large-scale data collected and sold to multiple clients (e.g., Nielsen,
GfK).
Used for tracking behavior like purchases, media consumption, etc.
Provides industry benchmarks but not tailored to one firm’s
problem.
GOOD TO UNDERSTAND
Why Analytics Is Booming
The key isn’t just knowing methods, it’s turning data → insight →
decision.
QUICK REVIEW / SELF-TEST
1. What’s the difference between a marketing decision problem and a
research problem?
2. Why does the level of measurement matter in data analysis?
3. How do exploratory, descriptive, and causal research designs differ?
4. Give one example of primary data and one of secondary data.
5. Why is analytics considered the “most desirable skill” for modern
marketers?
, Lecture 2: Measurement and Scaling:
Reliability, Validity, Dimensionality
MUST KNOW
Conceptual Model
Constructs/variables: Theoretical concepts we want to study (e.g.,
customer loyalty, price sensitivity).
Propositions/hypotheses: The expected relationships between
constructs, visualised as arrows (positive or negative).
Observable vs. Unobservable variables:
o Observable (manifest) → directly measurable (e.g., income,
sales, age).
o Unobservable (latent) → need indirect measurement (e.g.,
loyalty, attitude, satisfaction).
Measurement and Scaling
Measurement: Assigning numbers to objects based on rules.
Scaling: Creating a continuum along which objects are placed (e.g.,
attitude 1–5).
Reliability & Validity
Reliability: Consistency of measurement (does it give the same
result each time?) (when measured in the same condition).
o Types:
Test–retest (stability).
Internal consistency (Cronbach’s α).
Validity: Accuracy (does it measure what it’s supposed to?).
o A valid measure must be reliable, but a reliable measure is not
automatically valid.
Example:
The Schiphol restroom “smiley buttons” might be unreliable (people press
repeatedly) and invalid (not measuring cleanliness, but mood).
MUST KNOW
What Is Marketing Research? The systematic planning, collection,
analysis, and communication of data relevant to marketing decision
making.
Purpose:
To reduce uncertainty in managerial decisions.
To improve performance and profitability by supporting data-
driven insights.
Marketing Decision Problem vs. Marketing Research Problem
Marketing Decision Marketing Research Problem
Problem
Action-oriented (“What Information-oriented (“What do we need
should we do?”) to know?”)
Focuses on symptoms Focuses on underlying causes
Example: “Why are sales Example: “What factors influence
declining?” customer satisfaction?”
This distinction is essential for defining good research questions.
Einstein’s quote fits perfectly here: “The formulation of the problem is
often more essential than its solution.”
The Iceberg Principle
“What you see (the symptoms) is only the tip of the iceberg; the true
causes lie beneath the surface.”
Managers often react to visible issues (declining sales, poor
campaign performance) without identifying underlying causes
(brand perception, pricing, targeting).
Good marketing analytics digs deeper to uncover why.
,4. Types of Data (Quantitative vs. Qualitative)
Qualitative Data Quantitative Data
Non-numerical insights (text, Numerical data (can be measured
opinions, emotions) statistically)
Useful for exploring motives, Useful for measuring patterns,
emotions, ideas testing hypotheses
Examples: Focus groups, Examples: Surveys, experiments,
interviews, text analysis customer databases
Hybrid skills are increasingly important: combine intuition (qual) with
statistical evidence (quant).
Levels of Measurement (super important for data analysis)
Level Definition Example Statistical
Techniques
Nomin Categories with no order Gender Chi-square test
al (male/female)
Ordina Ordered categories, but Clothing size (S, Median, rank-
l no consistent spacing M, L) order tests
Interv Ordered, equal spacing, “Need for Mean,
al no true zero uniqueness” scale correlation
Ratio Equal intervals + true Company profit, Regression, t-
zero sales test
Why it matters:
The level of measurement determines which statistical tests you can apply.
Types of Marketing Research (by Research Design)
Type Goal Example Common
Techniques
Exploratory Gain insight, Focus groups, Qualitative
clarify ideas interviews, literature methods
review
Descriptive Describe “Do larger stores Surveys,
& relationships, test generate more regression
Predictive hypotheses sales?” analysis
Causal Test cause–effect “Does store size Experiments,
relationships cause higher sales?” ANOVA, t-test
(mediation and
moderation)
,Primary vs. Secondary Data
Primary Data Secondary Data
Collected for the current research Already collected for another
purpose
Examples: Surveys, interviews, Examples: Company reports,
experiments customer databases, industry data
More expensive but specific Cheaper but less tailored
Two types of secondary data:
Internal (sales records, CRM data)
and External (market reports,
syndicated data like Nielsen, GfK).
Syndicated Research
Large-scale data collected and sold to multiple clients (e.g., Nielsen,
GfK).
Used for tracking behavior like purchases, media consumption, etc.
Provides industry benchmarks but not tailored to one firm’s
problem.
GOOD TO UNDERSTAND
Why Analytics Is Booming
The key isn’t just knowing methods, it’s turning data → insight →
decision.
QUICK REVIEW / SELF-TEST
1. What’s the difference between a marketing decision problem and a
research problem?
2. Why does the level of measurement matter in data analysis?
3. How do exploratory, descriptive, and causal research designs differ?
4. Give one example of primary data and one of secondary data.
5. Why is analytics considered the “most desirable skill” for modern
marketers?
, Lecture 2: Measurement and Scaling:
Reliability, Validity, Dimensionality
MUST KNOW
Conceptual Model
Constructs/variables: Theoretical concepts we want to study (e.g.,
customer loyalty, price sensitivity).
Propositions/hypotheses: The expected relationships between
constructs, visualised as arrows (positive or negative).
Observable vs. Unobservable variables:
o Observable (manifest) → directly measurable (e.g., income,
sales, age).
o Unobservable (latent) → need indirect measurement (e.g.,
loyalty, attitude, satisfaction).
Measurement and Scaling
Measurement: Assigning numbers to objects based on rules.
Scaling: Creating a continuum along which objects are placed (e.g.,
attitude 1–5).
Reliability & Validity
Reliability: Consistency of measurement (does it give the same
result each time?) (when measured in the same condition).
o Types:
Test–retest (stability).
Internal consistency (Cronbach’s α).
Validity: Accuracy (does it measure what it’s supposed to?).
o A valid measure must be reliable, but a reliable measure is not
automatically valid.
Example:
The Schiphol restroom “smiley buttons” might be unreliable (people press
repeatedly) and invalid (not measuring cleanliness, but mood).
