Mixed Methods and Strategic Brand Applications

Introduction and Context

Brand decisions increasingly require two kinds of evidence. Quantitative research estimates incidence, market size, relationships, and change across populations. Qualitative research explains meanings, motivations, language, tensions, and context. Neither form of evidence is sufficient for every strategic question. A survey may show that perceived value predicts purchase intention, but not reveal what consumers mean by “value.” Interviews may reveal that consumers distrust a brand’s sustainability claims, but not establish how widespread that concern is or whether it affects market share.

Mixed-methods research deliberately integrates qualitative and quantitative approaches within one programme of inquiry. Integration is more than placing an interview study beside a survey. It involves a shared research problem, connected sampling and instruments, explicit sequencing or concurrency, and a reasoned process for combining findings. In brand management, mixed methods are especially useful because brands are simultaneously symbolic systems, commercial assets, customer experiences, and measurable market phenomena.

This chapter examines sequential and concurrent mixed-methods designs and applies them to segmentation, targeting, positioning, innovation, pricing, communications, customer experience, brand tracking, and portfolio decisions.

Learning Objectives

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

1. Explain the logic, designs, and quality criteria of mixed-methods research.
2. Distinguish exploratory, explanatory, embedded, and convergent designs.
3. Integrate qualitative and quantitative evidence without treating one as automatically superior.
4. Select an appropriate design for major brand-management decisions.
5. Evaluate validity, reliability, credibility, ethics, and managerial usefulness.
6. Translate integrated evidence into segmentation, positioning, innovation, pricing, communications, experience, tracking, and portfolio actions.

1. The Logic of Mixed Methods

A mixed-methods study combines qualitative and quantitative data in a single research project or connected programme. Its justification should be substantive: the combination must answer the research question better than either method alone.

Four rationales are particularly important:

• Complementarity: one method elaborates or clarifies findings from the other.
• Development: findings from one phase create instruments, hypotheses, stimuli, or sampling criteria for another.
• Expansion: different methods address different dimensions of a broad strategic problem.
• Triangulation: independently obtained evidence is compared to assess convergence, divergence, or contradiction.

Mixed methods should not be confused with methodological eclecticism. Using interviews, a survey, and social listening does not automatically create a mixed-methods design. The methods must be connected by a research purpose and an integration plan.

Paradigmatic foundations

Mixed-methods research is compatible with pragmatism, which prioritises the research question and the consequences of knowledge while allowing multiple forms of evidence. It can also be conducted from realist, critical, interpretivist, or post-positivist positions, provided the researcher makes assumptions explicit. A quantitative measure treats a construct as sufficiently standardised to compare across cases; a qualitative approach may treat meanings as situated and emergent. Integration therefore requires epistemic humility: numerical regularities and lived meanings are related, but not identical.

2. Principal Mixed-Methods Designs

Exploratory sequential design

The sequence is qualitative → quantitative. Researchers first explore consumer language, meanings, needs, or decision processes. They then use these findings to develop a questionnaire, scale, stimulus, segmentation variables, or hypotheses.

This design is appropriate when:

• the category is unfamiliar or changing;
• existing measures do not fit the market;
• the brand needs consumer-derived language;
• latent needs must be translated into measurable attributes.

Its main risk is reductionism: rich qualitative meanings may be compressed into overly narrow survey items.

Explanatory sequential design

The sequence is quantitative → qualitative. A survey, experiment, panel, or tracker identifies a pattern, after which qualitative research explains it.

For example, a brand tracker may show that awareness is high but consideration is low among younger consumers. Follow-up interviews can investigate whether the problem is irrelevance, distrust, price, poor availability, or a distinctive competitor advantage.

The qualitative sample is often purposive rather than representative. Participants may be selected from unusual, high-growth, declining, or contradictory segments revealed by the quantitative phase.

Convergent or concurrent design

Qualitative and quantitative data are collected during approximately the same period and integrated during interpretation. This is useful when a decision deadline is short or when the researcher wants parallel evidence about the same strategic issue.

For example, a national survey can estimate the prevalence of brand associations while online communities explore how those associations are formed. The designs may use different samples and instruments, but the constructs and decision criteria must be mapped before analysis.

Embedded design

One method has a primary role and the other is nested within it. A large concept test may embed open-ended interviews, ethnographic observation, or a small qualitative module. Alternatively, an ethnographic innovation project may embed a short survey to profile participants or quantify reactions to prototypes.

Multiphase design

Large brand programmes often contain several linked phases: exploratory interviews, a segmentation survey, qualitative work with priority segments, concept testing, and longitudinal tracking. The design is judged by the coherence of the chain, not by the number of methods used.

3. Designing Integration

Integration should be designed before fieldwork. A useful protocol includes five steps.

Step 1: Formulate an integrated research question

Instead of asking only “What is our brand awareness?”, ask: “How strong and widely distributed are our awareness and associations, and how do consumers interpret them in purchase contexts?” The first part invites measurement; the second invites interpretation.

Step 2: Map constructs across methods

Create a construct matrix identifying how each concept will be observed. For example, “trust” may be represented by spontaneous language in interviews, agreement with scale items in a survey, observed claim-checking behaviour, and repeat purchase in transaction data. These are related indicators, not interchangeable facts.

Step 3: Connect samples deliberately

A survey may use probability or quota sampling, whereas qualitative participants may be purposively selected. Connection can occur through shared screening variables, selection of survey respondents for interviews, or comparison of equivalent market cells. Researchers should document exclusions and recruitment effects.

Step 4: Analyse each dataset appropriately

Qualitative data may be coded thematically, abductively, narratively, or through framework analysis. Quantitative data may involve descriptive statistics, cross-tabulation, regression, factor analysis, clustering, conjoint analysis, or experimental modelling. Integration does not justify applying statistical tests to small purposive qualitative samples.

Step 5: Build a joint display

A joint display places findings side by side and records the inference that follows.

The final column must be a reasoned managerial inference, not a mechanical translation of a statistic.

4. Quality, Validity, and Interpretation

Quantitative quality commonly involves reliability, construct validity, statistical conclusion validity, and generalisability. Qualitative quality involves credibility, reflexivity, dependability, confirmability, and transferability. Mixed-methods quality adds integration quality: whether the methods genuinely address connected questions and whether conclusions are warranted by the combined evidence.

Contradiction is not automatically a failure. Suppose survey respondents rate ethical sourcing as important, while interviews reveal little spontaneous concern. Several explanations are possible: social desirability, low salience during ordinary choice, weak knowledge, or a gap between values and behaviour. The contradiction becomes a research finding requiring further investigation.

Researchers should distinguish four statements:

1. Description: what the data show.
2. Interpretation: what the pattern may mean.
3. Causal claim: what appears to produce change, requiring appropriate design.
4. Decision implication: what management should do, considering risk and resources.

A useful quantitative relationship

When a survey estimates a proportion, the standard error of a simple random-sample proportion is:

Here, p̂ is the sample proportion expressed from 0 to 1, and n is the sample size. The approximation assumes a simple random sample and is most appropriate when observations are independent. For a large-sample 95% confidence interval, a common approximation is:

The interval describes sampling uncertainty under the stated assumptions; it does not correct for non-probability sampling, measurement error, nonresponse bias, or poor questionnaire design. Mixed-methods evidence can illuminate those sources of uncertainty but does not eliminate them.

5. Applications Across Brand Strategy

Segmentation

Quantitative research can identify groups using needs, behaviours, attitudes, value, usage, or customer profitability. Common techniques include factor analysis followed by cluster analysis, latent class models, and predictive classification. Qualitative research helps determine whether segments are meaningful in lived experience, whether consumers recognise the proposed distinctions, and what language and occasions define them.

A statistically distinct cluster is not necessarily a useful segment. A managerial segment should be identifiable, substantial or strategically valuable, reachable, differentiable, and serviceable. Qualitative work can expose overlap between apparently separate clusters and reveal cultural or situational differences hidden by person-level averages.

Targeting

Targeting combines attractiveness with feasibility. Quantitative evidence estimates size, growth, spending, reach, and response. Qualitative evidence assesses unmet needs, identity fit, barriers, trust, and the social acceptability of a brand’s proposition. A small segment may deserve priority if it has high strategic value or acts as an influential early-adopter group, but this must be tested rather than romanticised.

Positioning

Positioning research should connect category frames, points of parity, points of difference, reasons to believe, and emotional or symbolic meaning. Qualitative methods generate and refine positioning territories. Quantitative methods assess distinctiveness, relevance, credibility, uniqueness, and purchase implications across competitors.

A perceptual map can be useful, but it is not a complete positioning strategy. Spatial proximity reflects selected variables and modelling assumptions; it does not prove causal advantage.

Innovation

Mixed methods support the innovation funnel:

1. Observe practices and frustrations.
2. Define opportunity spaces.
3. Co-create concepts.
4. Quantitatively screen concepts.
5. Prototype and test usage.
6. Forecast adoption and refine the offer.

Qualitative research is particularly valuable before consumers can articulate a solution. Quantitative concept testing is valuable when alternatives must be compared under standardised conditions. Researchers should avoid presenting a polished concept to participants and then attributing positive ratings solely to the underlying idea; execution effects and novelty effects may be substantial.

Pricing

Qualitative research identifies price meanings, fairness perceptions, reference prices, sacrifice, and the language consumers use to justify payment. Quantitative methods can estimate price sensitivity through controlled experiments, choice-based conjoint, discrete-choice experiments, or market tests.

In a simple linear demand model:

Q is quantity demanded, P is price, and a and b are estimated parameters with b positive in this formulation. The model is a simplification and assumes other relevant influences are held constant or included separately. Revenue is:

A price increase may raise or reduce revenue depending on the change in quantity. Survey intention alone is not a reliable forecast of actual demand; price experiments and behavioural data are generally stronger for commercial decisions.

Communications

Qualitative work develops consumer insight, creative territories, narrative codes, and potential sources of misunderstanding. Quantitative pretesting estimates attention, comprehension, brand linkage, persuasion, distinctiveness, and behavioural intention. Post-campaign tracking can assess reach, message take-out, brand effects, and sales-related outcomes.

A high liking score is not equivalent to effective advertising. A humorous execution may be remembered but poorly linked to the brand. Integration should therefore examine both meaning and measurable communication outcomes.

Customer experience

Experience research combines journey observation, diary studies, interviews, complaint analysis, operational metrics, satisfaction measures, and behavioural data. Qualitative evidence identifies emotionally significant moments and workarounds. Quantitative evidence identifies frequency, severity, segment differences, and links to retention or value.

A single satisfaction index can conceal severe failures in a low-frequency but high-consequence touchpoint. Journey-based mixed methods are therefore more informative than relying on one aggregate score.

Brand tracking

Tracking programmes benefit from stable quantitative measures over time, supplemented by periodic qualitative pulses. Quantitative trends show awareness, consideration, preference, associations, usage, and advocacy. Qualitative work explains emerging language, competitor narratives, cultural change, and reasons behind shifts.

Trend interpretation requires attention to questionnaire changes, sampling changes, seasonality, media exposure, market events, and statistical uncertainty. A movement of two percentage points may be strategically meaningful or merely sampling variation, depending on design and precision.

Portfolio decisions

Brand portfolios require evidence about cannibalisation, role clarity, price ladders, category coverage, architecture, and consumer navigation. Quantitative analysis can estimate overlap, switching, incremental reach, and profitability. Qualitative research explores confusion, perceived redundancy, endorsement effects, and the meanings attached to each brand.

A new sub-brand may attract existing customers rather than new demand. Conversely, a seemingly overlapping brand may serve a distinct usage occasion or channel. Portfolio decisions should therefore integrate market-level and consumer-level evidence with financial analysis.

6. Worked Application: The Northstar Meal Platform

The following is an illustrative case. Northstar, a national meal-kit brand, has high aided awareness but declining conversion among urban professionals aged 25–40. Management is considering a new flexible subscription and wants guidance on target market, proposition, pricing, and communication.

Background and problem

The existing tracker shows aided awareness of 78%, consideration of 31%, and recent trial of 12%. The figures suggest a conversion problem, but they do not explain whether the issue is price, effort, relevance, trust, or competitive substitution.

Phase 1: Explanatory quantitative study

A survey of category users measures usage occasions, perceived effort, price perceptions, cooking confidence, brand associations, trial, and intention. A regression model is used diagnostically to estimate association between consideration and predictors. The research team explicitly avoids interpreting association as proof of causality.

Results show that perceived effort and inflexibility are more strongly associated with non-consideration than absolute price. A small group reports high interest but very low confidence in completing recipes after work.

Phase 2: Qualitative explanation

Researchers recruit four purposive groups: high-intent non-triers, lapsed users, current heavy users, and users of a competitor. Mobile diaries and follow-up interviews reveal that “flexibility” means the ability to skip, pause, substitute, or order fewer meals—not merely a shorter contract. Participants also describe a fear of accumulating unused ingredients.

Phase 3: Integration and solution

The integrated inference is that Northstar’s conversion barrier is not simply expensive pricing. It is perceived commitment and waste risk. Management develops a proposition based on controllable commitment: one-off boxes, easy pauses, ingredient reuse, and recipes categorised by effort. A quantitative choice experiment compares subscription structures, delivery fees, recipe flexibility, and prices. Qualitative usability sessions then examine whether the winning configuration is understood.

Outcome and learning points

The launch decision uses four evidence layers: market incidence from the survey, meaning and barriers from diaries, preference trade-offs from the choice experiment, and comprehension from usability research. The case illustrates why a survey finding such as “price is important” should not be accepted as a complete diagnosis. Price may be a surface expression of risk, waste, or perceived value.

7. Technology, AI, and Data Governance

Digital platforms enable integrated research through behavioural logs, online communities, social listening, mobile diaries, dashboards, and rapid experimentation. AI can assist with transcription, translation, topic discovery, coding suggestions, text classification, and synthesis across large datasets. However, automated categories are not equivalent to valid constructs. Models may reproduce demographic, linguistic, or cultural bias and may miss irony, silence, power relations, and context.

Best practice requires human review, an audit trail, secure handling of personal data, transparent consent, and clear separation between exploratory pattern detection and confirmatory evidence. Researchers should disclose whether synthetic participants, automated respondents, or generative tools were used. Personal data should be minimised, purpose-limited, retained only as necessary, and handled in accordance with applicable privacy law and institutional standards.

8. Challenges and Limitations

Mixed methods are demanding. They require more time, integration expertise, budget, and project governance than a single method. Common failures include:

• collecting two datasets without a connecting question;
• allowing the largest sample to dominate interpretation;
• treating qualitative quotations as statistically representative;
• treating survey responses as transparent measures of truth;
• changing constructs between phases without documenting the change;
• using incompatible samples and then claiming direct comparison;
• resolving contradictions by discarding inconvenient evidence;
• confusing correlation with causal impact;
• reporting a sophisticated method without explaining the decision relevance.

Mixed methods also cannot guarantee truth. Poor sampling, weak measures, leading moderation, contaminated experiments, or selective reporting can affect every phase. Integration makes reasoning more robust only when the underlying studies are competent and their limitations are visible.

9. Best-Practice Decision Protocol

Before approving a mixed-methods programme, a brand team should be able to answer:

1. What decision will the research inform?
2. Which part of the problem requires measurement, explanation, or both?
3. Why is each method necessary?
4. How will findings be connected across samples, constructs, and time?
5. What pattern would count as convergence, complementarity, or contradiction?
6. What are the boundaries of inference?
7. How will uncertainty affect the decision?
8. What ethical, privacy, and representation risks exist?
9. Who owns the final interpretation and action plan?

The strongest research briefs specify not only deliverables, but also decision thresholds, alternative actions, and the cost of being wrong.

Chapter Summary

Mixed-methods research integrates qualitative and quantitative evidence to answer strategic brand questions more completely. Exploratory sequential designs develop measures from consumer meanings; explanatory sequential designs use qualitative work to explain quantitative patterns; convergent designs compare parallel evidence; embedded and multiphase designs support complex programmes. Integration requires connected questions, construct mapping, deliberate sampling, appropriate analysis, joint displays, and explicit inference.

In brand management, mixed methods improve segmentation and targeting by combining market structure with lived relevance; strengthen positioning by linking meaning to measurable distinctiveness; support innovation by moving from observation to testing; improve pricing decisions by combining price meanings with behavioural trade-offs; and enhance communications, experience, tracking, and portfolio decisions through explanation of trends and contradictions. The approach is powerful but not automatically valid. Quality depends on methodological competence, transparency, ethical governance, and disciplined interpretation.

Key Terms

Mixed methods: A research approach that intentionally integrates qualitative and quantitative evidence.

Triangulation: Comparison of evidence from different methods, sources, or researchers.

Complementarity: Use of one method to elaborate or clarify another.

Exploratory sequential design: Qualitative research followed by quantitative research.

Explanatory sequential design: Quantitative research followed by qualitative explanation.

Convergent design: Concurrent collection and integration of qualitative and quantitative evidence.

Embedded design: A secondary method nested within a primary research design.

Joint display: A structured presentation linking findings from different methods.

Construct: An abstract concept operationalised through observable indicators.

Credibility: Confidence that qualitative findings accurately represent the phenomenon studied.

Reliability: Consistency of a quantitative measurement process.

Validity: The extent to which evidence supports the intended interpretation of a measure.

Segmentation: Division of a market into meaningful and actionable groups.

Positioning: The intended distinctive meaning of a brand in a target customer’s mind.

Conjoint analysis: A quantitative method for estimating preferences among combinations of attributes.

Brand tracking: Repeated measurement of brand health indicators over time.

Cannibalisation: Loss of sales from an existing offering caused by another offering in the same portfolio.

Reflexivity: Critical examination of how the researcher influences research and interpretation.

Data governance: Policies and practices controlling data quality, access, privacy, security, and use.

Review Questions

Short Answer Questions

1. What distinguishes mixed-methods research from simply using several research techniques?
2. When is an exploratory sequential design most appropriate?
3. What is the purpose of a joint display?
4. Why can a statistically distinct segment still be strategically useless?
5. Name two ethical risks associated with AI-supported qualitative analysis.

Long Answer Questions

1. Compare exploratory sequential, explanatory sequential, and convergent designs. Explain the strategic situations in which each is appropriate.
2. Explain how mixed methods can improve brand positioning research, including the roles of qualitative insight and quantitative assessment.
3. Discuss how researchers should interpret contradictory qualitative and quantitative findings.
4. Evaluate the role of mixed methods in brand tracking and customer-experience management.
5. Explain the methodological and managerial limitations of mixed-methods research.

Analytical / Application Questions

1. A tracker reports high sustainability importance but no improvement in brand preference. Design an explanatory sequential study and specify the integration points.
2. A portfolio contains three brands with overlapping customer profiles. Propose a mixed-methods programme to assess cannibalisation, role clarity, and consumer confusion.
3. A new subscription concept receives strong survey intention but weak trial after launch. Diagnose the problem using quantitative and qualitative evidence, and recommend the next research phase.

Multiple Choice Questions with Answers

Question 1

Which design begins with qualitative research and uses its findings to develop a quantitative instrument?

A. Exploratory sequential design

B. Explanatory sequential design

C. Convergent design

D. Embedded design

Correct Answer: A. Exploratory sequential design

Question 2

A survey finds that consideration is low, and interviews are then conducted with high-awareness non-considerers to understand why. This is primarily an example of:

A. Exploratory sequential research

B. Explanatory sequential research

C. Pure ethnography

D. Experimental research only

Correct Answer: B. Explanatory sequential research

Question 3

What is the principal purpose of a joint display?

A. To replace sampling procedures

B. To present only statistically significant results

C. To connect findings from different methods and support integrated inference

D. To convert qualitative quotations into numerical data

Correct Answer: C. To connect findings from different methods and support integrated inference

Question 4

In the simple demand model Q = a − bP, what does b represent?

A. The intercept of the demand curve

B. The quantity demanded when price is zero

C. The estimated rate at which quantity changes as price changes, expressed as a positive magnitude in this formulation

D. Total revenue

Correct Answer: C. The estimated rate at which quantity changes as price changes, expressed as a positive magnitude in this formulation

Question 5

Which statement about a contradiction between survey and interview findings is most appropriate?

A. The qualitative findings must be discarded

B. The quantitative findings must be discarded

C. The contradiction may reveal issues such as social desirability, low salience, or a values-behaviour gap

D. Contradictory findings prove that the research is invalid

Correct Answer: C. The contradiction may reveal issues such as social desirability, low salience, or a values-behaviour gap

Question 6

Which evidence is most directly useful for determining whether a proposed segment is meaningful in lived experience?

A. Only the size of the segment

B. Qualitative exploration of needs, occasions, language, and barriers

C. Only aided awareness

D. The number of questionnaire items

Correct Answer: B. Qualitative exploration of needs, occasions, language, and barriers

Question 7

If p̂ = 0.50 and n = 100, what is the approximate standard error using SE(p̂) = √[p̂(1 − p̂)/n]?

A. 0.005

B. 0.05

C. 0.10

D. 0.50

Correct Answer: B. 0.05

Question 8

Why should a high advertising-liking score not be treated as proof of communication effectiveness?

A. Liking is always unrelated to advertising

B. An execution may be liked but poorly linked to the brand or behaviour

C. Quantitative measures cannot assess advertising

D. Qualitative research cannot study advertising

Correct Answer: B. An execution may be liked but poorly linked to the brand or behaviour

Question 9

Which is an appropriate use of qualitative research in pricing?

A. Estimating national demand without sampling

B. Identifying meanings of fairness, sacrifice, and reference price

C. Replacing all behavioural price experiments

D. Calculating revenue directly from interviews

Correct Answer: B. Identifying meanings of fairness, sacrifice, and reference price

Question 10

What is a central governance requirement when AI assists with qualitative coding?

A. Accepting automated categories without review

B. Removing the audit trail to protect speed

C. Human review, documentation, privacy protection, and attention to model bias

D. Using only synthetic participants

Correct Answer: C. Human review, documentation, privacy protection, and attention to model bias