Findings, Discussion and Managerial Decision-Making

Introduction

The findings chapter is the point at which a brand-management project becomes useful to both scholarship and practice. Data collection alone does not produce knowledge. The researcher must organise evidence, interpret patterns, relate results to theory and prior studies, acknowledge uncertainty, and translate conclusions into decisions that a brand team can implement.

A strong chapter distinguishes four activities:

1. Reporting what the evidence shows.
2. Interpreting what the evidence may mean.
3. Evaluating how credible, limited or transferable the evidence is.
4. Deciding what the organisation should do next.

These activities should be connected but not confused. A descriptive finding is not automatically a causal explanation, and a recommendation is not automatically justified merely because it appears attractive. The quality of managerial decision-making depends on the strength of the reasoning that connects evidence to action.

Learning Objectives

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

• present qualitative and quantitative findings clearly and selectively;
• distinguish evidence, inference, interpretation and speculation;
• relate findings to brand theories and previous research;
• evaluate validity, reliability, limitations and alternative explanations;
• convert insights into prioritised brand recommendations;
• design suitable KPIs, implementation stages and risk controls; and
• defend managerial decisions ethically and professionally.

1. The Logic of Evidence-Based Brand Management

Brand decisions commonly concern positioning, identity, communication, customer experience, portfolio architecture, distribution and investment. These decisions are made under uncertainty because customer responses are affected by competitors, culture, economic conditions, technology and changing expectations.

A useful reasoning chain is:

Evidence → finding → interpretation → implication → decision → measurement → learning

For example, a survey may show that younger customers are less likely to associate a brand with environmental responsibility than older customers. The finding is the observed difference. An interpretation might be that the brand’s sustainability communication is not reaching younger audiences or is not credible to them. A managerial implication could be to review content, media selection and proof points. The decision might involve a pilot campaign. Measurement then determines whether awareness, credibility and behaviour improve.

This chain prevents two common errors. The first is data dumping, in which tables are presented without meaning. The second is unsupported prescription, in which recommendations are made without showing how they follow from evidence.

1.1 Evidence, inference and speculation

These terms should be used precisely.

Language should reflect confidence. Use “the findings indicate” when evidence is reasonably consistent, “the results are consistent with” when an explanation is plausible, and “the data cannot establish” when causal or general claims exceed the design.

1.2 Descriptive, relational and causal claims

A descriptive study can report levels, frequencies or perceptions. A relational study can examine whether variables move together. A causal claim requires stronger evidence, normally including temporal order, a credible comparison and control of plausible alternative explanations.

For instance, if customers who follow a brand on social media report higher loyalty, the association does not prove that social-media following caused loyalty. Loyal customers may simply be more likely to follow the brand. A recommendation to increase social-media activity may still be reasonable, but it should be framed as a testable intervention rather than a proven causal solution.

2. Presenting Findings Clearly

2.1 Organise around research questions

Findings should normally follow the research questions or objectives rather than the order in which data were collected. Each section should answer one question and identify the evidence used. A practical structure is:

1. sample or participant profile;
2. findings for Research Question 1;
3. findings for Research Question 2;
4. integrated findings across methods; and
5. findings that directly inform decisions.

Background information should be included only when it helps the reader interpret the results. Demographic tables, for example, are useful when they explain segmentation or sample limitations, not when they merely increase length.

2.2 Quantitative presentation

Quantitative reporting should include the statistic, its meaning and its relevance. Percentages require a denominator. “Forty per cent agreed” is incomplete unless the reader knows whether this means 40% of 50 respondents or 40% of 5,000 customers.

For a proportion, the basic calculation is:

Here, p is the sample proportion, x is the number of respondents with the characteristic, and n is the total number of valid observations. The calculation assumes that n is clearly defined and that missing responses are handled consistently. If 84 of 200 respondents recognise a brand, p = 84 ÷ 200 = 0.42, or 42%.

When comparing groups, report both the size and direction of the difference. If recognition is 42% among new customers and 63% among existing customers, the percentage-point difference is 21 points. This is not the same as saying recognition is 21% higher. The relative increase is:

The percentage-point difference is often easier for managers to interpret, while the relative increase can exaggerate practical importance when the baseline is small.

For business funnels, conversion rate may be calculated as:

The numerator and denominator must refer to the same period, audience and definition of action. A campaign conversion rate should not mix impressions with unique eligible users unless the research design explicitly justifies it.

A recommendation based on survey findings should also consider uncertainty, sample quality and effect size. Statistical significance, where calculated, does not by itself establish managerial importance. A small difference may be statistically detectable but commercially irrelevant; a meaningful difference may fail to reach significance in a small sample.

2.3 Qualitative presentation

Qualitative findings should be analytical rather than anecdotal. Direct quotations can demonstrate a theme, but one memorable quote is not evidence that all participants shared the view. A good thematic presentation identifies:

• the theme;
• its scope and variation;
• representative evidence;
• exceptions or negative cases; and
• its implication for the research question.

For example, a theme called “convenience as reassurance” might show that customers valued fast delivery not only for speed but because it reduced perceived purchase risk. Contrasting participants who prioritised price would prevent the analysis from treating the theme as universal.

2.4 Integrated or mixed-method findings

When qualitative and quantitative methods are combined, integration should occur at the level of meaning. A survey may show low trust, while interviews explain that unclear product claims cause the distrust. The two forms of evidence are complementary: the survey indicates prevalence and the interviews provide interpretation.

Results may also diverge. Divergence should be investigated, not concealed. Respondents may state that sustainability matters in a survey but choose primarily on price in observed behaviour. Such a result may indicate a gap between stated values and purchase constraints, social desirability bias, or differences between intention and behaviour.

3. Discussion: Connecting Findings to Theory and Research

The discussion answers the question: “So what does this evidence mean in relation to what is already known?” It should not repeat the findings section. Instead, it should compare, explain and qualify.

3.1 Relevant brand frameworks

Several frameworks may support interpretation:

• Brand equity: consumer-based brand equity concerns the value created by awareness, associations, perceived quality and loyalty. Findings should specify which dimension is affected rather than describing all positive responses as “equity”.
• Brand identity and image: identity concerns what the organisation seeks to project; image concerns how stakeholders perceive the brand. A gap between intended identity and perceived image can explain communication or positioning problems.
• Positioning: positioning concerns the distinctive and credible meaning a brand seeks to occupy relative to alternatives. Findings should test distinctiveness, relevance, credibility and consistency.
• Customer-based brand resonance: strong relationships may involve salience, performance, imagery, judgements, feelings and resonance. The framework can help identify whether a problem is awareness-based or relationship-based.
• Service-dominant logic: value is often co-created through interactions rather than delivered solely by the firm. This is useful when findings concern employees, platforms, communities or customer participation.

Theory should guide interpretation, not be used as decoration. A project need not mention every model encountered in the literature. It should use the framework that best explains the research problem and acknowledge where the findings do not fit.

3.2 Confirming, extending and challenging prior research

A finding may:

• confirm an established relationship in a new setting;
• extend prior work by identifying a new mechanism, audience or context;
• qualify a relationship by showing that it applies only to some segments; or
• challenge prior research through contradictory evidence.

Contradiction does not automatically mean that one study is wrong. Differences may result from sample composition, measurement, timing, market maturity, cultural context or research method. The discussion should offer plausible explanations while distinguishing explanation from proof.

3.3 Alternative explanations

Good discussion tests its preferred interpretation against alternatives. If awareness is low, possible explanations include weak media reach, confusing category structure, limited distribution, recent market entry or measurement error. If loyalty is low, possible explanations include price sensitivity, switching incentives, poor service recovery or strong competitor performance.

A useful paragraph structure is:

1. state the principal interpretation;
2. identify the evidence supporting it;
3. relate it to theory or prior research;
4. acknowledge a plausible alternative; and
5. explain what additional evidence would discriminate between them.

4. Evaluating Research Quality and Limitations

Limitations are not an apology. They define the boundaries within which conclusions are credible.

4.1 Common limitations

Sampling limitations arise when the sample is small, non-random, geographically narrow or dominated by one customer group. Findings may be useful for the studied context without being statistically generalisable to the whole market.

Measurement limitations occur when concepts such as trust, loyalty or authenticity are represented by weak, ambiguous or leading questions. A single item may provide a useful indication but rarely captures a complex construct completely.

Method limitations include cross-sectional designs, self-report bias, interviewer effects and lack of behavioural data. Cross-sectional evidence captures a situation at one point in time and is weak for demonstrating change or causality.

Analytical limitations include selective coding, multiple comparisons, overinterpretation of subgroups and failure to account for missing data. Transparency about analytical decisions strengthens credibility.

Contextual limitations include unusual market conditions, competitor actions, economic shocks or a campaign occurring during data collection.

4.2 Validity, reliability and credibility

Validity concerns whether the study supports the intended interpretation. Reliability concerns consistency of measurement. In qualitative work, credibility may be strengthened through transparent coding, reflexivity, triangulation, negative-case analysis and participant clarification where appropriate.

Limitations should be connected to consequences. “The sample was convenient” is incomplete. A stronger statement is: “Because participants were recruited through the brand’s online community, highly engaged customers may be overrepresented; therefore, the findings should not be used to estimate the attitudes of inactive or lapsed customers without further research.”

4.3 Prioritising limitations

Not every limitation has equal importance. Prioritise limitations that could change:

• the direction of a conclusion;
• the size of an estimated effect;
• the suitability of a target segment; or
• the ranking of recommended actions.

A limitation that affects generalisability may not prevent a carefully designed pilot. Conversely, uncertainty about whether a problem exists at all should normally be resolved before major investment.

5. From Insight to Managerial Decision

An insight becomes decision-useful when it identifies a problem or opportunity, explains why it matters, indicates who is affected, and suggests an action that can be tested.

5.1 The recommendation test

Each recommendation should answer six questions:

1. What should be done?
2. Why is it needed?
3. Which evidence supports it?
4. Who owns it?
5. How will success be measured?
6. What could go wrong?

Recommendations should be specific enough to implement but not more certain than the evidence permits. “Improve the brand” is not actionable. “Test clearer sustainability proof points on product pages for environmentally concerned first-time buyers” is more precise and testable.

5.2 Prioritising recommendations

A simple prioritisation matrix can compare expected impact with feasibility. Impact may include revenue, retention, brand equity, customer experience or risk reduction. Feasibility may include cost, capability, time, legal constraints and operational complexity.

A weighted scoring model may be used transparently. For example:

Here, I is expected impact, F is feasibility, C is confidence in the evidence, R is risk, and w₁, w₂, w₃, w₄ are weights that sum to 1 if the components are measured on comparable scales. The formula is a decision aid, not an objective truth. Scores are meaningful only when criteria and scales are defined consistently.

5.3 KPIs and measurement design

KPIs should connect activity to an outcome. A communications activity metric, such as reach, is not equivalent to a brand outcome such as consideration or a commercial outcome such as repeat purchase.

A balanced measurement set may include:

• Reach: target audience exposure;
• Brand: awareness, associations, perceived quality, trust or consideration;
• Behaviour: trial, conversion, repeat purchase, retention or advocacy;
• Experience: satisfaction, complaint resolution or task completion;
• Financial: margin, customer lifetime value or return on investment.

Return on investment can be expressed as:

“Incremental” means the return attributable to the activity relative to a credible baseline or comparison. The formula assumes that return and investment are measured on the same basis and period. Revenue should not be substituted for profit unless the organisation explicitly defines the measure as revenue return.

Each KPI needs a baseline, target, owner, data source, measurement frequency and decision rule. A pilot might specify: “If qualified consideration increases by at least five percentage points without reducing conversion or increasing complaint rates, expand the activity; otherwise revise or stop it.”

5.4 Implementation implications

Implementation should be staged:

1. Prepare: clarify proposition, owners, resources, compliance checks and baseline measures.
2. Pilot: implement with a defined segment, channel or region.
3. Evaluate: compare results with baseline or control where possible.
4. Scale: expand only when evidence supports wider deployment.
5. Institutionalise: update processes, training, dashboards and governance.

This approach converts uncertainty into managed learning rather than pretending that a report can eliminate uncertainty.

6. Ethical, Governance and Technology Considerations

Brand recommendations affect people, not just metrics. Segmentation may unintentionally exclude vulnerable groups or encourage discriminatory targeting. Personalisation can become intrusive when customers do not understand how data are used. Sustainability claims require evidence and should avoid vague or exaggerated language. Influencer and AI-generated content should be disclosed where required and reviewed for accuracy, bias, copyright and brand-safety risks.

AI can assist with coding, sentiment classification, scenario generation and dashboard monitoring, but it should not replace researcher judgement. Automated outputs may reproduce bias, misclassify irony or conceal uncertainty. Sensitive data should be minimised, protected and processed according to applicable privacy requirements. Human review, audit trails and documented decision rules are essential.

7. Illustrative Case: Repositioning a Sustainable Personal-Care Brand

Background

This fictional case concerns “Luma,” a mid-sized personal-care brand that claims to offer effective products with lower environmental impact. Sales growth has slowed, although aided awareness remains high.

Problem

Management wants to know whether the problem is weak awareness, low credibility, poor value perception or an inconsistent customer experience. It is considering a major packaging redesign and a large media investment.

Analysis

A mixed-method project surveys 400 category buyers, interviews 20 customers and reviews online complaints. The survey finds that 74% recognise Luma, but only 31% spontaneously name it when asked about sustainable personal-care brands. Among those aware of the brand, 48% consider its environmental claims credible. Interviews suggest that customers value refillability but cannot easily verify disposal instructions. Complaints show that refill components are sometimes unavailable.

The evidence indicates that awareness is not the principal constraint. The more defensible interpretation is a credibility and availability problem. However, the cross-sectional design cannot prove that improving disposal information will increase repeat purchase. Price and distribution may be alternative explanations.

Approach and solution

The recommended sequence is:

1. improve product-page and pack information with specific, verifiable claims;
2. correct refill availability in priority retailers;
3. pilot the revised proposition in two regions;
4. measure credibility, consideration, refill purchase and complaints; and
5. delay a full identity redesign until the pilot identifies whether visual change is necessary.

The proposal is prioritised because it addresses high-confidence problems with lower cost and reversibility than a complete redesign. Key risks include greenwashing allegations, increased packaging complexity and stock-outs. Legal and sustainability specialists should approve claims before release.

Outcome and learning points

In the pilot, management would compare results with a matched region or pre-pilot baseline. Success should not be defined by impressions alone. It should include a meaningful increase in verified-claim credibility, improved refill availability, stable or improved conversion, and fewer disposal-related complaints.

The case demonstrates three principles: high awareness does not guarantee strong brand meaning; customer experience can undermine positioning; and staged experimentation is preferable to an irreversible investment when diagnosis remains uncertain.

8. Best Practices for the Final Chapter

A professionally written findings and discussion chapter should:

• use headings that mirror the research questions;
• report relevant evidence before interpreting it;
• distinguish association from causation;
• include contradictory and negative evidence;
• connect claims to theory without forcing theoretical language;
• state limitations and their practical consequences;
• rank recommendations rather than presenting an undifferentiated list;
• assign owners, resources, timelines and decision rules;
• use KPIs that measure outcomes as well as activity; and
• state what further research or testing is required.

The final test is whether a reader can understand not only what the researcher discovered, but also why the proposed action is justified, how it can be implemented, and how the organisation will know whether it worked.

Chapter Summary

Findings, discussion and managerial decision-making form a connected process. Findings present patterns in the evidence; discussion interprets those patterns through theory, prior research and alternative explanations; and recommendations translate credible insights into action. Effective analysis distinguishes descriptive, relational and causal claims and uses appropriately cautious language.

A strong chapter reports quantitative and qualitative evidence transparently, integrates methods where appropriate, and treats limitations as boundaries on interpretation. Recommendations should be specific, evidence-linked, prioritised and measurable. KPIs require baselines, targets, owners, data sources and decision rules. Pilots, comparison groups and staged implementation help organisations learn under uncertainty. Ethical governance is necessary when decisions involve personal data, vulnerable audiences, sustainability claims or AI-supported analysis.

Key Terms

Evidence: Information produced through a systematic research process.

Finding: A concise statement summarising a relevant evidence pattern.

Inference: A reasoned interpretation supported by evidence.

Speculation: An untested or weakly supported possible explanation.

Causality: A relationship in which one factor contributes to producing change in another.

Association: A relationship in which variables occur together without proof of causation.

Brand equity: Value associated with a brand’s awareness, meanings, perceived quality and relationships.

Brand identity: The meaning and associations an organisation seeks to establish.

Brand image: Stakeholder perceptions of the brand.

Positioning: The distinctive and credible place a brand seeks to occupy in customers’ minds.

Triangulation: Comparing evidence from different sources or methods.

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

Reliability: The consistency of a measurement or procedure.

Generalisability: The extent to which findings apply beyond the studied sample or context.

KPI: A key performance indicator used to monitor progress toward an objective.

Baseline: The starting measurement against which change is assessed.

Pilot: A limited implementation used to test an intervention.

Incrementality: The change attributable to an activity beyond what would otherwise have occurred.

Governance: Structures and processes that guide accountability, compliance and responsible decision-making.

Review Questions

5 Short Answer Questions

1. Distinguish a finding from an inference.
2. Why should percentage points be distinguished from relative percentage change?
3. What is the purpose of relating findings to theory and prior research?
4. Identify three common limitations in brand-management research.
5. What information should accompany a KPI?

5 Long Answer Questions

1. Explain how a researcher can distinguish descriptive, relational and causal claims in a findings chapter.
2. Discuss how qualitative and quantitative evidence can be integrated when the two methods produce different results.
3. Evaluate the role of brand-equity and positioning frameworks in interpreting research findings.
4. Explain why limitations should be connected to their practical consequences rather than merely listed.
5. Describe a staged process for converting a research insight into an implemented brand strategy.

3 Analytical / Application Questions

1. A survey shows that customers who follow a brand on social media have higher loyalty scores. Analyse two alternative explanations and design a suitable follow-up study.
2. A brand has high aided awareness but low consideration and poor repeat purchase. Develop three prioritised recommendations, including evidence requirements, KPIs and risks.
3. Construct a decision rule for a pilot sustainability campaign. Specify the baseline, target measures, comparison approach and conditions for scaling or stopping.

Multiple Choice Questions with Answers

Question 1

Which statement best distinguishes evidence from inference?

A. Evidence is always causal, whereas inference is always descriptive.

B. Evidence is systematically produced information, whereas inference is a reasoned interpretation of that information.

C. Evidence is managerial opinion, whereas inference is a statistical calculation.

D. Evidence is speculation supported by theory, whereas inference is an untested assumption.

Correct Answer: B. Evidence is systematically produced information, whereas inference is a reasoned interpretation of that information.

Question 2

A study finds that 84 of 200 respondents recognise a brand. What is the sample proportion?

A. 21%

B. 32%

C. 42%

D. 84%

Correct Answer: C. 42%

Question 3

Why does a positive association between social-media following and loyalty not prove causation?

A. The sample percentage cannot be calculated.

B. Loyal customers may be more likely to follow the brand.

C. Social media cannot be used in brand management.

D. Loyalty is never measurable.

Correct Answer: B. Loyal customers may be more likely to follow the brand.

Question 4

Which practice most improves the analytical quality of qualitative findings?

A. Presenting only the most memorable quotation

B. Reporting themes together with variation and negative cases

C. Removing quotations from the report

D. Treating every participant statement as equally prevalent

Correct Answer: B. Reporting themes together with variation and negative cases

Question 5

What is the main purpose of a baseline?

A. To replace the research question

B. To provide a starting point for assessing change

C. To guarantee causal validity

D. To identify the brand logo

Correct Answer: B. To provide a starting point for assessing change

Question 6

Which recommendation is most actionable?

A. Improve the brand image.

B. Increase marketing activity everywhere.

C. Test clearer sustainability proof points on product pages for first-time buyers.

D. Make customers more loyal.

Correct Answer: C. Test clearer sustainability proof points on product pages for first-time buyers.

Question 7

In the formula ROI = (Incremental return − Investment) ÷ Investment × 100, what does incremental return mean?

A. Total market revenue regardless of the activity

B. Revenue or profit that would have occurred without the activity

C. The return attributable to the activity beyond a credible baseline or comparison

D. The investment divided by the number of impressions

Correct Answer: C. The return attributable to the activity beyond a credible baseline or comparison

Question 8

Why might a full brand redesign be deferred in the Luma case?

A. Awareness was already high and the more immediate problems concerned credibility and availability.

B. Packaging can never affect brand meaning.

C. Research cannot inform design decisions.

D. Refill products are unrelated to sustainability positioning.

Correct Answer: A. Awareness was already high and the more immediate problems concerned credibility and availability.

Question 9

Which KPI set best combines brand and commercial outcomes?

A. Impressions only

B. Logo size, colour preference and staff opinions

C. Credibility, consideration, repeat purchase and complaint rate

D. Number of meetings and pages in the report

Correct Answer: C. Credibility, consideration, repeat purchase and complaint rate

Question 10

What is a key governance concern when AI is used to analyse customer sentiment?

A. AI always produces causal evidence.

B. Automated systems may reproduce bias or misclassify context and therefore require human review.

C. AI eliminates the need for data protection.

D. AI results cannot be documented.

Correct Answer: B. Automated systems may reproduce bias or misclassify context and therefore require human review.