Research Problems, Questions and Insight Briefs

Introduction

Market research rarely fails because researchers cannot collect data. It more often fails because the organization has not clearly defined what decision the evidence must support. A vague request such as “understand declining brand performance” may lead to an attractive but unusable report. Effective research begins earlier: with the translation of a managerial concern into a decision problem, research objectives, information requirements, questions, hypotheses, scope, governance arrangements and success criteria.

For brand managers, this translation is particularly important. Brands are intangible, socially constructed and experienced across multiple touchpoints. A decline in sales may reflect weak awareness, poor distribution, changing category needs, price perceptions, competitive activity, communication wear-out or an internal measurement problem. Each explanation implies a different research design and a different managerial response.

This chapter presents a disciplined process for developing research problems, research questions and insight briefs. It combines qualitative and quantitative reasoning and shows how a research brief can connect organizational decisions to credible evidence.

Learning objectives

By the end of this chapter, readers should be able to:

1. Distinguish a managerial decision problem from a research problem.
2. Convert a broad brand challenge into research objectives and information requirements.
3. Formulate clear qualitative questions and testable quantitative hypotheses.
4. Select an appropriate scope, stakeholder structure, budget logic and timeline.
5. Construct an insight brief that guides research without prejudging the findings.
6. Evaluate the quality, feasibility, ethics and usefulness of a proposed research brief.

1. From brand challenge to decision problem

A brand challenge is an observed or suspected business difficulty, opportunity or uncertainty. Examples include falling consideration, low trial among a new segment, inconsistent brand meaning across markets, weak retention or uncertainty about a proposed repositioning.

A decision problem specifies what a decision-maker may need to decide. It is action-oriented and normally includes alternatives. For example:

Should the brand invest in a new value proposition for young professionals, improve distribution, revise its pricing architecture, or retain the current strategy?

A research problem defines the knowledge gap that prevents the decision from being made confidently. It may be expressed as:

The organization does not know whether low purchase among young professionals is primarily caused by weak relevance, limited availability, price-value concerns or ineffective communication.

The distinction matters because research cannot directly solve every business problem. Research produces evidence; management makes choices using that evidence alongside financial, operational, legal and strategic considerations.

The problem translation chain

A useful chain is:

Business symptom → decision problem → research problem → objectives → questions or hypotheses → evidence → insight → decision

For example:

A strong problem statement is specific enough to guide research but open enough to permit unexpected explanations.

2. Decision problems and research problems

Decision problems are usually framed in terms of action, risk and resource allocation. Research problems are framed in terms of uncertainty, concepts and evidence. Confusing them creates two common errors.

First, a team may commission research before identifying the decision. It then collects broad descriptive data that nobody is prepared to use. Second, the team may define the research problem as a preferred answer, such as “prove that the new campaign will work.” This introduces confirmation bias and undermines the credibility of the study.

A useful decision-problem statement identifies:

• the decision owner;
• the decision to be made;
• the realistic alternatives;
• the consequences of being wrong;
• the decision date; and
• the evidence required to reduce uncertainty.

A useful research-problem statement identifies:

• the population or market context;
• the phenomenon to be understood or measured;
• relevant constructs and relationships;
• the boundaries of the investigation; and
• what will remain outside the study.

Problem diagnosis questions

Before writing a brief, the researcher should ask:

1. What has changed, and compared with what baseline?
2. Is the issue a symptom, a cause or an interpretation of data?
3. Which decision will the research inform?
4. What options are genuinely available to management?
5. What is already known from internal and external evidence?
6. What uncertainty is consequential enough to investigate?
7. What evidence would change the decision?

These questions help distinguish a researchable uncertainty from a general desire for more information.

3. Research objectives and information requirements

Research objectives state what the study will accomplish. They should use observable verbs such as identify, explore, describe, measure, compare, explain, evaluate and estimate. Objectives should not use vague verbs such as “look at” or “understand” without specifying the object of understanding.

A well-formed objective usually includes a population, phenomenon and purpose:

To measure perceived value, convenience and trust among current and lapsed users of the category, and to determine how these perceptions differ by age and usage frequency.

Objectives should be prioritized. A study with twelve objectives may have no real priority and may exceed its budget or timetable. One distinction is between:

• Primary objectives, which are necessary for the decision;
• Secondary objectives, which improve interpretation; and
• Out-of-scope questions, which may be interesting but are not required now.

Information requirements

Information requirements are the specific evidence needed to address each objective. They may include:

• category behaviours and purchase occasions;
• awareness, salience and consideration;
• brand associations and perceived differentiation;
• functional and emotional needs;
• price and value perceptions;
• channel access and availability;
• reactions to concepts, packaging or communications;
• barriers to trial, use or retention; and
• segment size, profile and commercial potential.

An information-requirements matrix prevents the brief from becoming a list of disconnected questions.

4. Research questions and hypotheses

A research question is an explicit question that directs evidence collection and analysis. It should be answerable with the proposed design. Research questions may be exploratory, descriptive, explanatory, evaluative or predictive.

Qualitative questions

Qualitative questions investigate meanings, experiences, processes, language and context. They are often open-ended:

• How do consumers define a “reliable” solution in this category?
• What occasions make the category relevant or irrelevant?
• How do people explain switching from the brand?
• What tensions exist between convenience and perceived quality?

Qualitative questions should not force respondents to adopt the researcher's concepts. Asking “Why do you dislike our premium positioning?” assumes both dislike and the relevance of premium positioning. A less leading formulation is: “How would you describe the brand’s value in relation to alternatives?”

Quantitative questions

Quantitative questions specify measurable variables, populations and comparisons:

• What proportion of category buyers recognize the brand?
• How does consideration differ between current users and non-users?
• Is perceived convenience associated with purchase intention after controlling for price sensitivity?
• Which segment has the highest combination of need, preference and accessibility?

A quantitative question should clarify the unit of analysis, such as person, household, purchase occasion, advertisement or market.

Hypotheses

A hypothesis is a logically derived, empirically testable proposition. It is not simply a prediction or a desired result. A directional hypothesis might be:

H₁: Among category users, perceived convenience is positively associated with purchase intention.

The corresponding null hypothesis is:

H₀: In the target population, perceived convenience is not positively associated with purchase intention.

The wording must match the design. A cross-sectional survey can identify association, but normally cannot establish that convenience perceptions cause purchase intention. Causal language requires stronger conditions, such as temporal precedence, control of alternative explanations and, ideally, experimental or quasi-experimental evidence.

For a simple linear relationship, a researcher may specify:

Here, Y is the outcome, such as purchase intention; X is a predictor, such as perceived convenience; β₀ is the intercept; β₁ is the expected change in Y for a one-unit change in X; and ε represents unexplained variation. This model assumes an appropriate scale interpretation, a correctly specified relationship and independent errors for ordinary least-squares estimation. It describes association unless the research design supports causal inference.

Hypotheses should be derived from theory, prior evidence or a clearly articulated managerial logic. A long list of post hoc hypotheses increases the risk of chance findings and selective reporting.

5. Building an insight brief

An insight brief is a decision-oriented document that aligns sponsors, researchers and users of evidence. It is more than a procurement request. It explains why the study matters, what must be learned, how findings will be used and what constraints govern the work.

Recommended structure

#### 5.1 Background and context

Describe the brand, category, market development, relevant performance evidence and strategic context. Separate verified facts from assumptions. “Consideration fell from 38% to 31% in the tracking study” is different from “the campaign caused the decline.”

#### 5.2 Business decision

State the decision, decision owner, alternatives and decision date. The brief should identify whether the research will support a strategic choice, product development, communication refinement, segmentation, channel planning or performance diagnosis.

#### 5.3 Research problem and objectives

Present the central uncertainty and a prioritized set of objectives. Include what the research will not address.

#### 5.4 Target population and scope

Define geography, category eligibility, customer status, age or other relevant characteristics. Clarify whether the study concerns users, non-users, lapsed users, trade partners, employees or multiple stakeholder groups. Specify time period, channels, brands and markets.

#### 5.5 Proposed approach

The sponsor may indicate desired methods, but should avoid prescribing a method before the problem is understood. A mixed-method design may combine qualitative exploration with quantitative measurement. For example, interviews can reveal the language and mechanisms behind a problem; a survey can estimate prevalence and compare segments.

#### 5.6 Outputs and decisions

Specify the outputs: a diagnostic report, segment profiles, concept territories, dashboard, workshop or decision recommendation. Outputs should be linked to decisions rather than merely to deliverables.

#### 5.7 Stakeholders and governance

Name the sponsor, decision owner, research lead, data owner, agency or supplier, legal or ethics reviewer and intended users. Establish who approves the questionnaire, who controls access to data and who signs off the interpretation.

#### 5.8 Budget and timeline

Budget should reflect sample, recruitment, incentives, fieldwork, translation, analysis, technology, travel, data protection and contingency. A timeline should include briefing, design, pilot, recruitment, fieldwork, cleaning, analysis, reporting and decision workshops. Compressing the schedule without reducing scope may threaten quality.

#### 5.9 Success criteria

Success is not simply delivery on time. Criteria may include:

• decision-maker confidence in the evidence;
• completion of all primary objectives;
• acceptable data quality and sample coverage;
• actionable differentiation among strategic options;
• stakeholder adoption of the findings; and
• measurable movement in a subsequent business or brand indicator, where a valid follow-up period exists.

Success criteria should distinguish research quality from business outcomes. A sound study may reveal that a proposed strategy is unattractive; that is a valuable result even if it does not produce an immediate performance increase.

6. Qualitative and quantitative alignment

Qualitative and quantitative research are not interchangeable. Qualitative work is especially useful when the problem is poorly understood, language is uncertain, meanings are important or the researcher needs to examine social and contextual processes. Quantitative work is useful when constructs can be operationalized and the organization needs estimates, comparisons, relationships or projections.

A sequential exploratory design may proceed as follows:

1. Conduct interviews or groups to discover needs, vocabulary and hypotheses.
2. Translate themes into measurable constructs and response options.
3. Pilot the questionnaire.
4. Conduct a survey to estimate incidence and segment differences.
5. Integrate findings, examining convergence and meaningful divergence.

A sequential explanatory design reverses the emphasis: quantitative results identify an unexpected pattern, and qualitative work explains it.

Integration should occur at the level of the decision, not merely by placing two reports side by side. If a survey indicates low consideration but interviews reveal strong admiration, the apparent contradiction may reflect a distinction between symbolic respect and practical relevance. The contradiction is an insight to investigate, not a reason to discard one method.

7. Stakeholder management and scope control

Research projects involve different forms of expertise. Senior sponsors understand strategic consequences; brand teams understand market history; analysts understand measurement; legal teams understand consent and data obligations; agencies understand fieldwork. Their interests may conflict.

A stakeholder map can classify participants by influence and involvement:

Scope creep often begins with apparently small additions: another market, a new segment, a competitor module or an extra wave. Each addition affects sample, questionnaire length, analysis and timetable. Change control should ask whether the addition affects a primary decision, what it costs and what must be removed to accommodate it.

8. Budget, timeline and research value

A research budget should be built from activities rather than from a single arbitrary total. Typical cost categories include design, recruitment, respondent incentives, fieldwork, sample access, translation, transcription, data processing, analysis, reporting, technology and contingency.

The value of research depends on the decision context. A useful conceptual model is expected value of information:

Here, EVI is expected value of information; a represents possible actions; U(a) is the utility or value of an action; and the two terms compare the expected value when information is available with the best expected value before obtaining it. This is a conceptual decision-analysis expression, not a guaranteed financial forecast. It assumes that possible states, actions and utilities can be represented credibly. In practice, uncertainty about probabilities and outcomes means that EVI should be treated as a structured way to discuss value, not as false precision.

A timeline should protect the stages that most affect validity: problem definition, instrument design, piloting, recruitment and data quality checks. Reporting time should not be eliminated merely to meet a launch date. If the decision date is fixed, the scope or method may need adjustment.

9. Ethics, governance and responsible insight

Research briefs create ethical obligations before data collection begins. Participants should receive understandable information about purpose, procedures, risks, incentives, confidentiality and withdrawal. Consent should be appropriate to the population and method. Special care is required with children, vulnerable groups, sensitive topics and identifiable behavioural data.

Researchers should collect only data necessary for the stated purpose, restrict access, define retention periods and use secure transfer and storage. Personally identifiable information should be separated from research data where possible. Applicable privacy and data-protection law must be considered across jurisdictions.

The sponsor must not use research as disguised sales activity. Participants should not be misled about the organization or the use of their responses. Incentives should compensate time without becoming coercive.

Responsible interpretation requires avoiding stigmatizing segment labels, unsupported causal claims, selective reporting and fabricated precision. AI tools may assist with transcription, coding, questionnaire review or synthesis, but outputs require human validation. Sensitive data should not be entered into an unapproved system. AI-generated themes must be traceable to source material, and automated classification should be checked for language, cultural and demographic bias.

10. Quality criteria for a research brief

A high-quality brief is:

• Decision-linked: it identifies the choice the evidence will inform.
• Specific: it defines population, constructs, context and boundaries.
• Neutral: it does not assume the preferred answer.
• Feasible: its objectives fit the budget, sample and timeline.
• Methodologically coherent: questions, measures and analysis are aligned.
• Actionable: findings can distinguish among realistic options.
• Ethically responsible: participant rights and data governance are explicit.
• Transparent: assumptions, limitations and success criteria are visible.

A brief should be reviewed by asking whether a different research supplier would reach the same understanding of the problem. If not, ambiguity remains.

Applied case: Reframing the “young consumer” problem

Background

The fictional brand NaturaSip, a premium ready-to-drink tea, has strong awareness but declining trial among consumers aged 21–30. The marketing team believes that young consumers perceive the brand as “old-fashioned.” The team requests focus groups to test a new youthful advertising campaign.

Problem diagnosis

The request contains an assumed cause and a preferred solution. Sales data show that the decline is concentrated in convenience stores, while supermarket performance is stable. Existing tracking shows high trust but moderate relevance. The real decision is whether NaturaSip should change its communication, reformulate or repackage the product, expand convenience distribution, or pursue a combination.

The research problem is therefore:

NaturaSip lacks evidence about the relative contribution of relevance, occasion fit, price-value perceptions, product cues and convenience-store availability to low trial among young adult category buyers.

Objectives

1. Explore how young adult buyers define desirable ready-to-drink tea occasions.
2. Identify the meanings attached to NaturaSip and competing brands.
3. Measure the prevalence of barriers to trial and repeat purchase.
4. Compare barriers between convenience-store shoppers and supermarket shoppers.
5. Evaluate whether alternative positioning territories improve perceived relevance without weakening trust.

Approach

The proposed design begins with mobile ethnographic diaries and individual interviews to capture consumption occasions and shopping context. Findings are used to develop a survey among category buyers aged 21–30, with quotas for channel and usage frequency. A concept test then compares positioning territories.

The survey measures awareness, consideration, recent purchase, perceived relevance, convenience, value, trust, distinctiveness, availability and purchase intention. It does not claim that any single perception causes purchase intention because the design is cross-sectional.

Illustrative outcome

Qualitative evidence reveals that respondents do not reject the brand as old-fashioned. Instead, they associate it with planned supermarket purchases and afternoon relaxation, whereas convenience-store shoppers seek an immediately visible, cold and energizing option. Quantitative results show that perceived availability and occasion fit are more common barriers than lack of trust.

Decision and learning

NaturaSip prioritizes convenience-channel visibility, cold availability and an “on-the-go refreshment” usage occasion while retaining its trust credentials. The case demonstrates that a research brief should test competing explanations rather than simply validate the first managerial interpretation. It also shows the value of integrating contextual qualitative evidence with quantitative estimation.

Challenges and limitations

Problem definition is affected by organizational politics. Sponsors may deliberately keep objectives broad to preserve flexibility, or they may narrow them to protect an existing strategy. Researchers must be tactful but explicit about assumptions.

Research questions can also become overambitious. A single study cannot usually diagnose a market, create a segmentation, forecast sales, test every communication and explain long-term loyalty at the same time. Prioritization is a methodological requirement, not merely a project-management preference.

Qualitative findings are rich but not automatically representative. Quantitative findings may be generalizable within a well-defined sample but can still be misleading if constructs are poorly measured, respondents are inattentive or the sample excludes relevant groups. Mixed methods improve understanding only when the methods are coherently designed and genuinely integrated.

Finally, insight is not identical to a finding. “Thirty-two percent selected option A” is a finding. An insight explains why that pattern matters, how it connects to behaviour or strategy and what action it enables. Insight must remain grounded in evidence; it is not permission for unsupported storytelling.

Chapter Summary

Effective market research begins by translating a brand challenge into a decision problem and then into a research problem. The decision problem concerns action; the research problem concerns the uncertainty that evidence can reduce. Clear objectives, information requirements, research questions and hypotheses create alignment between the decision and the study design.

Qualitative questions explore meanings, experiences and mechanisms, whereas quantitative questions estimate prevalence, compare groups or examine relationships. Hypotheses must be testable and should not imply causality beyond the design. An insight brief should define context, decision, objectives, scope, methods, stakeholders, governance, budget, timeline, outputs and success criteria. It must also address ethics, privacy, responsible AI use and limitations.

The central professional discipline is prioritization: collect the evidence required for a consequential decision, not information merely because it is available. A strong brief is neutral, feasible, methodologically coherent and actionable.

Key Terms

Decision problem: The managerial choice or action that research may inform.

Research problem: The knowledge uncertainty that prevents a decision from being made confidently.

Brand challenge: An observed or suspected business difficulty or opportunity involving a brand.

Research objective: A specific statement of what the study will accomplish.

Information requirement: Evidence needed to address an objective or decision issue.

Research question: An explicit, answerable question guiding evidence collection and analysis.

Hypothesis: A testable proposition about a phenomenon or relationship.

Null hypothesis: A proposition representing no specified effect or relationship.

Construct: An abstract concept, such as trust or relevance, that must be operationalized.

Unit of analysis: The entity being studied, such as a person, household, occasion or market.

Qualitative research: Research examining meanings, experiences, processes and context, usually through non-numerical data.

Quantitative research: Research using numerical measurement to estimate, compare or model phenomena.

Mixed methods: Deliberate integration of qualitative and quantitative approaches.

Insight brief: A decision-oriented document aligning context, objectives, methods, resources and intended use.

Scope: The defined boundaries of a research project.

Stakeholder: A person or group affected by, contributing to or using the research.

Expected value of information: The potential improvement in decision value resulting from additional information.

Research governance: Structures controlling accountability, ethics, data access and decision rights.

Scope creep: Uncontrolled expansion of project requirements after agreement.

Operationalization: Translating an abstract construct into observable measures or indicators.

Review Questions

Short Answer Questions

1. Distinguish between a managerial decision problem and a research problem.
2. What makes a research objective well formed?
3. Give two examples of non-leading qualitative research questions.
4. Why should a cross-sectional survey normally avoid strong causal claims?
5. Identify four components that should appear in an insight brief.

Long Answer Questions

1. Explain the translation chain from a brand challenge to a managerial decision, using an original example.
2. Compare the roles of qualitative and quantitative research in a sequential mixed-method design.
3. Discuss how scope, budget and timeline influence research quality.
4. Explain how hypotheses should be developed and how their wording must reflect the research design.
5. Evaluate the ethical and governance requirements that should be addressed before collecting consumer data.

Analytical / Application Questions

1. A brand has high awareness but low consideration among first-time category buyers. Construct a decision problem, research problem, three objectives and four information requirements.
2. A sponsor asks for focus groups “to prove” that a new campaign is better. Critique the request and rewrite it as a neutral, decision-linked brief.
3. Design an integrated qualitative and quantitative approach for diagnosing declining repeat purchase. Explain the sequence, stakeholders, success criteria and principal limitations.

Multiple Choice Questions with Answers

Question 1

Which statement best describes a research problem?

A. The organization should increase its advertising budget.

B. The organization lacks evidence about the factors explaining declining preference.

C. The brand manager prefers a younger visual identity.

D. The research agency must deliver a report by Friday.

Correct Answer: B. The organization lacks evidence about the factors explaining declining preference.

Question 2

Which objective is most appropriately written for a quantitative study?

A. Look at what consumers think about the brand.

B. Understand everything about the category.

C. Measure perceived value and compare it across current and lapsed users.

D. Prove that the campaign will succeed.

Correct Answer: C. Measure perceived value and compare it across current and lapsed users.

Question 3

Which question is least leading in qualitative research?

A. Why do you dislike our premium positioning?

B. How would you describe the brand’s value compared with alternatives?

C. What makes our new campaign persuasive?

D. Why is our packaging too traditional?

Correct Answer: B. How would you describe the brand’s value compared with alternatives?

Question 4

A cross-sectional survey finds that perceived convenience and purchase intention are correlated. What conclusion is most defensible?

A. Convenience definitely causes purchase intention.

B. Purchase intention definitely causes convenience perceptions.

C. The variables are associated, but causality is not established by the design alone.

D. Neither variable can be measured quantitatively.

Correct Answer: C. The variables are associated, but causality is not established by the design alone.

Question 5

What is the primary purpose of an information-requirements matrix?

A. To replace the research report.

B. To connect decision issues with needed evidence and possible methods.

C. To guarantee a statistically representative sample.

D. To calculate the agency’s profit margin.

Correct Answer: B. To connect decision issues with needed evidence and possible methods.

Question 6

Which item is most clearly a decision-problem statement?

A. Consumers use several competing brands.

B. The brand has moderate trust scores.

C. Management must decide whether to change positioning or increase channel investment.

D. The questionnaire contains twelve measures.

Correct Answer: C. Management must decide whether to change positioning or increase channel investment.

Question 7

In the expression Y = β₀ + β₁X + ε, what does ε represent?

A. The intercept

B. The predictor variable

C. The unexplained component or error term

D. The sample size

Correct Answer: C. The unexplained component or error term

Question 8

Which practice best demonstrates responsible use of AI in market research?

A. Accepting automated themes without checking source data

B. Uploading identifiable participant data to any free online tool

C. Using AI for preliminary coding while validating outputs and protecting sensitive data

D. Allowing AI to determine the final managerial decision

Correct Answer: C. Using AI for preliminary coding while validating outputs and protecting sensitive data

Question 9

Which is an appropriate success criterion for a research project?

A. The findings confirm the sponsor’s preferred strategy.

B. The study distinguishes among realistic strategic options with acceptable data quality.

C. The report contains the greatest possible number of charts.

D. All interesting questions are included, regardless of budget.

Correct Answer: B. The study distinguishes among realistic strategic options with acceptable data quality.

Question 10

Why should qualitative and quantitative findings be integrated at the decision level?

A. Because qualitative research is always more accurate.

B. Because two reports are never useful separately.

C. Because integration connects meanings and mechanisms with estimates and comparisons relevant to action.

D. Because quantitative research cannot measure consumer attitudes.

Correct Answer: C. Because integration connects meanings and mechanisms with estimates and comparisons relevant to action.