Untitled

• Marketing research projects involve defining a problem and formulating a research design.

Triggers for Research

• Research is often needed when evaluating alternatives. This includes determining the right message for a new product (e.g., a new iPhone).
• Opportunities, such as emerging technologies or government incentives for electric vehicles (EVs), can also trigger research. Companies should seize these opportunities and research them.
• Threats, like government regulations (e.g., a TikTok ban), require research to find alternative platforms for advertising.
• New competitors entering the market, such as BlueSky affecting X (formerly Twitter), necessitate research to understand user migration and retention strategies.

Defining the Problem

• It involves translating management decision problems into marketing research problems.
• Management decision problems (e.g., declining sales) require actionable solutions and focus on symptoms, asking "what should the decision-maker do?".
• Marketing research problems are information-oriented, focusing on root causes and asking "where can I find the info?".

Formulating the Research Design

• There is no single ideal research design.
• Exploratory (qualitative) and conclusive (quantitative) research designs both have pros and cons.
• Descriptive research aims to understand the state of a product or company by describing "how many, when, etc.".
• Casual research explores cause-and-effect relationships.

Exploratory vs. Conclusive Research

• Exploratory research is open-ended, uses small samples, unstructured data, and non-statistical analysis. It is broad and open-ended, providing deeper insights into motivations, attitudes, and beliefs.
• Exploratory research helps to understand consumers. It goes beyond functional benefits and helps in brand positioning and ideation.
• Conclusive research is not described.

Focus Groups: Origins

• Focus groups originated as a way to sell the war to the American public more effectively by changing their opinions.
• Ernest Dichter introduced using the same procedure to sell products after the war, becoming the father of focus groups in marketing.

Focus Groups: Accomplishments

• Focus groups help in understanding how users use a product, gather feedback, and improve the product.
• They can be a product of qualitative research, as seen with the Google Digital Wellbeing App.

Brand Management Using Focus Groups

• Focus groups can identify new segments of brand users and understand what drives brand loyalty.
• Reebok used research to understand that people liked the “vintage” look of their shoes, which helped them make targeted ads.

Designing Focus Groups

• Participants should be knowledgeable about the topic (6-10 people).
• Avoid professional research participants; seek homogeneity within groups and heterogeneity across groups. People might conform to others' decisions or want to speak on behalf of their demographic.
• The environment should be quiet, comfortable, and allow for recording during a 1-2 hour duration. The process should be carefully moderated.

Focus Groups: Pros

• Focus groups gather a higher volume of information from several people at once, and can be faster than in-depth interviews.
• Focus groups provide a better understanding of people's actual reactions. They are more convincing and impactful than quantitative data.

Focus Groups: Cons

• Groupthink and conforming is possible due to group setting.
• There is a risk of biases from the moderator who might ask leading questions.
• Focus groups may produce noisy data, and the views expressed may not represent the whole market.
• There is a temptation to dismiss any negative feedback.

In-Depth Interview (IDI)

• An IDI is a one-on-one conversation:
• IDIs aim to understand deeply what a person thinks about a product. Useful when discussing sensitive topics.
• The laddering technique helps uncover true motivations by going from concrete product traits to abstract benefits.

IDI Projective Techniques

• Projective techniques indirectly assess motivations and attitudes, useful when people are unaware, unwilling to share, or think it's irrelevant.
• Word association involves saying the first thing that comes to mind when a list is read.
• Sentence completion gives partial questions for people to fill in.
• Personification tries to understand brand image, which may show a unified or non-unified perception.

When NOT to Conduct Research

• Avoid research when there is disagreement about the problem, results are meaningless, resources are insufficient, or costs outweigh benefits.
• Also not conduct research if the opportunity has passed, decisions are already made, or needed information already exists.

Primary vs. Secondary Data

• Primary data is newly collected for a specific purpose. Used in exploratory, descriptive, and casual research, it’s tailored, actionable, and accurate.
• Secondary data is existing data gathered for a different purpose, used in exploratory and descriptive research, like Statista and Mintel.

Advantages of Secondary Data

• Advantages: faster and cheaper than primary collection methods. It also permits more flexibility, and has a wide-range of free sources.

Disadvantages of Secondary Data

• Disadvantages: often insufficient, ambiguous, and outdated, and may not address the specific research question.

Evaluating Secondary Data

• Two main questions must be answered: How was the data created? How can it be used?
• To know how it was created: understand the original study's purpose, sponsorship, methodology, sampling, and timing.
• To know how it can be used: ask how relevant it is to the original question and the company involved (same sector, competitor, different market.)

Tools

• Statista: Provides industry and brand-level data in simple tables that summarize complex information.
• Social Listening: Monitoring social media channels (Reddit, TikTok, X, Instagram) to understand what consumers are saying about a brand, product, competitors, or industry.
• Google Trends: Useful for comparing brands, assessing interest over time, identifying seasonal trends, and analyzing regional differences.
• YouTube: Brand YouTube Channels and consumer generated videos
• Reddit: Unfiltered and organic reviews of products

Research Designs

• Descriptive research aims to describe and answers well-defined questions through conclusive and quantitative methods. It does not establish causality, using common methods like surveys.

Measurements

• Measuring involves assigning numbers or categories to concepts of interest, allowing for statistical analysis and comparisons.

Steps in the Measurement Process

• Conceptualize: Define what you want to measure.
• Operationalize: Determine what you will observe.
• Measure: Specific survey questions.
• Evaluate: Asses the sources of measurement error.

Levels of Measurement

• Nominal Scale (lowest level): Distinguishes one from another, like listing streaming apps used.
• Ordinal Scale: Ranks preferences from high to low.
• Interval Scale: Rates disagreement with a statement.
• Ratio Scale: Measures something you can count.

Four Levels of Measurement

• Nominal: Can calculate percentages and the mode, e.g., which streaming apps do you use?
• Ordinal: Values can be ranked, and percentages, the mode, and the median can be calculated.
• Interval: The distance between values is constant, allowing for calculating percentages, the mode, the median, and standard deviation.
• Ratio: Has a meaningful zero, where percentages, mode, median, mean, and standard deviation can be calculated.

Ranking vs. Rating Scales

• Ranking puts things in order using an ordinal scale. Rating uses an interval or near-interval scale.

Random vs. Systematic Error

• Any measurement has: the true score, random error (uncontrollable), and systematic error (controllable).

Two Types of Systematic Error

• Measurement Error: Biases due to flaws in the actual measurement process.
• Sample Design Error: Biases due to flaws in the sampling design.

5 Measurement Errors

• Surrogate Information Error: Not measuring what you're supposed to be measuring.
• Measurement Instrument Bias: Problems with survey design and questions, e.g., leading questions.
• Nonresponse Bias: Participants who don't respond are systematically different than those who do respond.
• Response Bias: Tendency to respond inaccurately, consciously or subconsciously.
• Interviewers are influenced by respondents - giving an incentive for a desire response

Tests for Validity

• Face validity: Asks if you measure what you're set out to measure in a purely subjective sense.
• Content validity: Asks if the measure represents the subject in it's entirety.
• Predictive validity: Asks can the measure forecast future outcomes
• Convergent validity: Measure correlates strongly with other measures of the same construct
• Discriminant validity: Measure correlates weakly with other measures of unrelated constructs

Validity and Reliability

• Validity: Asks more about accuracy and how close the measurement is to the true score.
• Reliability: Asks more about consistency and precision. Asks is the measurement consistent across situations, respondents, and time

Survey Flaws

• 1-3 3-5 5-6 questions are BAD
• Response options are not mutually exclusive exclusive
• Demographics are grouped together
• Questions start with specific then have a general ending.