How to build a marketing research course: a complete guide for lecturers

Central question: How do we turn a managerial decision into a researchable problem without asking the study to decide the strategy for us?

Marketing research starts before a questionnaire, interview guide or dataset exists. The first teaching task is to separate the management decision problem from the research problem, then translate uncertainty into objectives and questions that can be answered with evidence. This distinction prevents students from gathering large amounts of data that are interesting but unusable.

What this concept covers

• Management decision problems versus marketing research problems.
• Research objectives, research questions and information needs.
• Decision criteria, assumptions and evidence gaps.
• Exploratory, descriptive and causal questions.
• Research briefs, scope, constraints, timing and stakeholder expectations.

Learning outcomes

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

• Formulate a research problem from a business decision.
• Write specific research objectives and answerable questions.
• Identify which uncertainty can and cannot be resolved through research.
• Specify the evidence required before a recommendation can be made.
• Draft a concise research brief that aligns the research with the decision.

Supports Course Outcomes 1 and 2.

How to teach it

• Open with a deliberately vague management request such as "find out why sales are down" and ask students what decision is actually pending.
• Require groups to rewrite the request as a management decision, research problem and three research objectives.
• Make students identify one attractive question they would explicitly exclude because it will not change the decision.
• Use short briefs from different industries so students learn that the logic transfers across products, services and B2B contexts.

Case-style example

Aurora Meals has seen monthly subscription renewals fall from 74% to 66% while acquisition remains stable. Management can spend £120,000 either improving onboarding, lowering price by 8%, or adding family plans. Students must write the decision problem, three research questions and the evidence needed for each option. One group may prioritise churn drivers; another may first test whether the apparent decline is concentrated in one cohort. Students discover that a research problem is a decision-focused specification, not a topic.

Where students struggle

Students frequently jump from a business symptom to a preferred method. They suggest "run a survey" before clarifying the decision, the population, the relevant comparison or what evidence would change management action.

Reading and a quick check

Use Nunan, Birks and Malhotra, Chapters 1-2. Quick check: give students a one-sentence management request and ask them to state the decision problem, research problem and one piece of information that is not worth collecting.

Applied teaching activity

Use a short case, research-design workshop, evidence critique or analyst memo so students leave the concept with a visible decision output rather than only lecture notes.

Where it leads

Once the research problem is defensible, students can design an evidence strategy that matches the question, resources and required confidence.

Central question: What combination of exploratory, descriptive and causal evidence is strong enough for the decision, given time, budget and uncertainty?

Research design is the architecture of the study. Lecturers should teach students to choose methods because of the inference required, not because a familiar tool is available. A strong design states what each source will contribute, how different methods connect, what validity threats remain and where the evidence will stop short of a causal conclusion.

What this concept covers

• Exploratory, descriptive and causal designs.
• Cross-sectional, longitudinal and experimental logic.
• Qualitative, quantitative and mixed-method designs.
• Triangulation, validity, reliability and generalisability.
• Research proposals, budgets, timelines and decision thresholds.

Learning outcomes

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

• Match research questions to appropriate designs.
• Explain the inference each method can and cannot support.
• Build a mixed-method design with a clear sequence and purpose.
• Identify major threats to validity and reliability.
• Defend research trade-offs under budget and time constraints.

Supports Course Outcomes 1 and 2.

How to teach it

• Give every group the same decision but different budgets and deadlines so design choices cannot be copied.
• Ask students to draw the evidence chain from question to method to sample to analysis to decision.
• Require a "what this design cannot tell us" box in every proposal.
• Use peer review in which one group acts as the client challenging another group's design.

Case-style example

BrightWave Appliances has £75,000 and eight weeks to decide whether to launch a £179 smart kettle. Management wants category demand, reasons for non-adoption and a price response estimate. Students allocate the budget across desk research, 20 interviews, a 600-person survey and an optional 2 x 2 price-message experiment. They must justify the sequence and cut one method if the budget falls to £45,000. Students discover that good design is a portfolio of evidence choices, not a maximal list of methods.

Where students struggle

Students often equate methodological sophistication with quality. They need practice recognising that a simpler design can be more decision-useful when the research question, sample and inference are tightly aligned.

Reading and a quick check

Use Nunan, Birks and Malhotra, Chapter 3, and McGivern, Chapters 4-7. Quick check: ask students which design they would choose to estimate incidence, explore motivation and establish causality, and why the same method should not answer all three.

Applied teaching activity

Use a short case, research-design workshop, evidence critique or analyst memo so students leave the concept with a visible decision output rather than only lecture notes.

Where it leads

The next step is to determine what is already known before commissioning primary research.

Central question: What can existing evidence tell us, how trustworthy is it, and where does it leave a material evidence gap?

Secondary research is not a preliminary internet search. It is a structured evaluation of existing internal and external evidence, with attention to definitions, dates, populations, collection methods and commercial incentives. Teaching it well helps students avoid duplicating research and prepares them to challenge market-size claims, trend narratives and competitor evidence.

What this concept covers

• Internal records, syndicated data, public statistics, industry reports and digital sources.
• Source provenance, recency, definitions and methodological fit.
• Market sizing and triangulation.
• Competitive and macro-environmental evidence.
• Evidence matrices, source logs and reproducible desk-research trails.

Learning outcomes

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

• Locate relevant secondary evidence efficiently.
• Evaluate source quality and comparability.
• Triangulate conflicting estimates rather than choosing the most convenient number.
• Build a defensible market-size range.
• Identify what primary research must still resolve.

Supports Course Outcomes 2 and 3.

How to teach it

• Give teams three conflicting market estimates and require them to audit definitions before calculating an average.
• Have students create a source-quality rubric covering provenance, population, collection method, date and incentives.
• Ask groups to show the full chain from a headline statistic to the original source.
• Use PESTLE only after students can distinguish evidence from framework labels.

Case-style example

CloudTrail Mobility is considering three UK cities. One industry report estimates the addressable market at £1.8bn, a consultancy estimates £2.6bn, and public licence data imply £1.3bn under a narrower definition. Students reconcile definitions, score each source on a five-point quality rubric, produce a market-size range and identify two gaps requiring primary research. One defensible view uses the public baseline; another uses the wider category after adjustment. Students discover that triangulation is an argument about definitions and evidence quality.

Where students struggle

Students often cite the newest or largest number without checking whether it measures the same market, population or behaviour. They also confuse a framework such as PESTLE with evidence itself.

Reading and a quick check

Use Nunan, Birks and Malhotra, Chapters 4-5. Quick check: ask students to explain why two credible sources can disagree without either being "wrong."

Where a simulation fits

PESTLE Analysis Simulation: After this concept, the PESTLE Analysis Simulation can turn desk evidence into a weighted macro-environmental judgement. It works best when students are already able to question source quality, definitions and materiality rather than merely fill six boxes.

Where it leads

Secondary evidence identifies what is known and what remains uncertain. Qualitative research then helps students explore meanings, motivations and unexpected explanations.

Central question: How can researchers uncover motivations, meanings and mechanisms without pretending that small qualitative samples estimate population prevalence?

Qualitative research is most useful when the researcher needs depth, language, context or theory-building rather than incidence. Students should learn to design interviews and groups, probe without leading, recognise researcher effects, code evidence systematically and distinguish a compelling theme from a population estimate.

What this concept covers

• Depth interviews, focus groups, observation, diaries, online communities and projective approaches.
• Recruitment, discussion guides and probing.
• Saturation, reflexivity and moderator effects.
• Coding, thematic analysis and evidence displays.
• Using qualitative insight to develop hypotheses, measures and decision options.

Learning outcomes

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

• Choose an appropriate qualitative method for an exploratory objective.
• Write a neutral discussion guide with productive probes.
• Conduct and critique an interview or focus group.
• Code qualitative material into themes with an audit trail.
• Translate themes into research hypotheses without over-generalising.

Supports Course Outcomes 3 and 4.

How to teach it

• Demonstrate the difference between a leading question and a neutral probe.
• Run a short interview in class and have observers code the same transcript independently.
• Ask students to distinguish theme frequency from theme importance.
• Require quotations to be linked to a coding frame and decision implication rather than used decoratively.

Case-style example

LumaSkin is testing a refillable moisturiser. Four focus groups of eight consumers, 12 depth interviews and 2,000 product reviews produce three recurring themes: hygiene anxiety, refill inconvenience and environmental identity. Students code 24 supplied excerpts, compare inter-coder disagreements and propose five survey items to test the themes quantitatively. One team may treat hygiene as a barrier; another may see it as a packaging-design opportunity. Students discover that qualitative evidence generates mechanisms and language, not population percentages.

Where students struggle

Students commonly report "60% of interviewees said..." or treat the loudest focus-group participant as representative. They need repeated practice separating exploratory meaning from statistical estimation.

Reading and a quick check

Use Nunan, Birks and Malhotra, Chapters 6-9. Quick check: ask students what a qualitative finding can legitimately support and what requires a quantitative follow-up.

Applied teaching activity

Use a short case, research-design workshop, evidence critique or analyst memo so students leave the concept with a visible decision output rather than only lecture notes.

Where it leads

Exploratory insight supplies constructs and language. The next concept turns those ideas into measures that can be administered consistently.

Central question: How do we convert an abstract construct into questions and scales that respondents can answer and researchers can interpret?

Measurement is where many apparently quantitative projects fail. Students need to connect constructs to operational definitions, choose scale formats, write unbiased items, manage order effects and pilot instruments before fieldwork. The teaching emphasis should be on validity and respondent cognition, not on cosmetic survey design.

What this concept covers

• Construct definition and operationalisation.
• Nominal, ordinal, interval-like and ratio measures.
• Likert-type, semantic differential and comparative scales.
• Question wording, order, routing and response options.
• Pilot testing, reliability and validity evidence.

Learning outcomes

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

• Define constructs before writing items.
• Select scale formats appropriate to the variable and analysis.
• Identify leading, double-barrelled, ambiguous and loaded questions.
• Design a coherent questionnaire with routing and respondent flow.
• Plan a pilot and specify what evidence would trigger revision.

Supports Course Outcomes 4 and 5.

How to teach it

• Give students a flawed questionnaire and make them diagnose each error before rewriting it.
• Use cognitive interviewing: one student answers while explaining what they think each item means.
• Have groups measure the same construct using different scales and compare information loss.
• Require every instrument to include a short construct-to-item matrix.

Case-style example

Nevo Fitness wants to measure satisfaction, perceived value and renewal intention. Its draft survey has 27 items, 11 reverse-worded statements, a double-barrelled "easy and motivating" item and a 1-10 scale whose endpoints change across pages. Students define the three constructs, cut the instrument to 18 items and design a 30-person pilot with revision rules. One view favours shorter scales; another protects content validity. Students discover that questionnaire length and measurement quality create real trade-offs.

Where students struggle

Students often begin with wording rather than construct definition, assume more items always mean better reliability, or treat a familiar Likert scale as automatically valid.

Reading and a quick check

Assign Hulland, Sample and Houston (2026) on scale development and Hulland, Baumgartner and Smith (2018) on survey-research best practice. Quick check: ask students to repair three flawed items and state the construct each is intended to measure.

Applied teaching activity

Use a short case, research-design workshop, evidence critique or analyst memo so students leave the concept with a visible decision output rather than only lecture notes.

Where it leads

Once the instrument is defensible, students need a sampling and fieldwork plan that tells them whose answers they are actually collecting.

Central question: Who must be observed, how will they be recruited, and what makes the achieved sample credible for the decision?

Sampling connects the target population to the data actually analysed. A rigorous course should distinguish probability and non-probability designs, expose coverage and nonresponse risks, teach sample-size logic without formula worship and address modern online-panel quality issues. Fieldwork should be taught as controlled data production, not an administrative afterthought.

What this concept covers

• Target population, sampling frame and sampling unit.
• Probability and non-probability sampling.
• Stratification, clustering, quotas and weighting.
• Sample size, precision, subgroup requirements and power.
• Online panels, recruitment, attention checks, fraud and fieldwork monitoring.

Learning outcomes

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

• Define a target population and assess frame coverage.
• Choose and justify a sampling design.
• Estimate an appropriate sample size for the decision.
• Identify likely selection, nonresponse and panel-quality biases.
• Create a fieldwork monitoring plan and transparent limitations statement.

Supports Course Outcomes 5 and 6.

How to teach it

• Ask students to compare a customer-list sample with a general-population panel for the same decision.
• Give subgroups with different minimum n requirements so a single headline sample size is insufficient.
• Use a recruitment funnel to show how eligibility and nonresponse change the achieved sample.
• Have students write the limitations paragraph before fieldwork rather than after results disappoint.

Case-style example

Kora Coffee has 48,000 loyalty members but wants insight about all 350,000 adults in its catchment. The budget allows 1,200 completes, including at least 200 lapsed buyers and 250 non-customers. Students design a sample using the loyalty list plus an external panel, set quotas, decide whether to weight the final data and calculate the effective implications of subgroup oversampling. One view maximises representativeness; another protects decision-critical subgroups. Students discover that the "sample size" is a design choice, not one number.

Where students struggle

Students often believe a large convenience sample is representative by virtue of size. They also underestimate coverage error, panel quality and the instability introduced by aggressive weighting.

Reading and a quick check

Assign Moss et al. (2023) on market-research panels. Quick check: ask why 5,000 self-selected responses can be less useful than 800 well-sampled responses.

Applied teaching activity

Use a short case, research-design workshop, evidence critique or analyst memo so students leave the concept with a visible decision output rather than only lecture notes.

Where it leads

Sampling defines the evidential boundary. Students can now prepare, summarise and visualise data without losing sight of uncertainty and data quality.

Central question: What does the dataset actually say before we run sophisticated models, and how should uncertainty and data quality shape the story?

Descriptive analysis is the point at which researchers learn whether their data are fit for purpose. Students should inspect missingness, distributions, outliers, coding, cross-tabs and uncertainty before choosing advanced methods. Visualisation should clarify a decision, not hide weak measurement behind attractive dashboards.

What this concept covers

• Data cleaning, coding, missing data and outliers.
• Frequency distributions, cross-tabulations and summary statistics.
• Confidence intervals and practical uncertainty.
• Weighting and subgroup comparisons.
• Decision-oriented charts, tables and dashboard critique.

Learning outcomes

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

• Prepare a raw dataset for analysis transparently.
• Select descriptive summaries that match the measurement scale.
• Interpret cross-tabs and confidence intervals.
• Distinguish statistical variation from managerial importance.
• Design visual displays that support rather than distort a research conclusion.

Supports Course Outcomes 6 and 7.

How to teach it

• Give students a deliberately messy dataset and require a reproducible cleaning log.
• Show two charts using different axis choices and ask which interpretation changes.
• Require a one-sentence decision implication beneath every table.
• Make students identify one result that should not be reported because the subgroup base is too small.

Case-style example

Streamly has 1,200 survey responses on subscription video use. Twelve percent of income values are missing, one region is oversampled two-to-one, and a 28-person subgroup shows 71% purchase intent versus 52% overall. Students create a cleaning log, weighted cross-tabs and confidence-aware visuals, then decide whether to feature the subgroup. One view sees a promising niche; another rejects the estimate as unstable. Students discover that descriptive analysis is also a judgement about data quality.

Where students struggle

Students often treat cleaning choices as invisible housekeeping and report every percentage as equally precise. They also confuse visual prominence with evidential strength.

Reading and a quick check

Use Nunan, Birks and Malhotra, Chapters 19-20. Quick check: ask students which three diagnostics they would inspect before any inferential model and why.

Applied teaching activity

Use a short case, research-design workshop, evidence critique or analyst memo so students leave the concept with a visible decision output rather than only lecture notes.

Where it leads

With a trustworthy descriptive picture, students can test hypotheses and design experiments that support stronger claims.

Central question: When can we say an intervention caused a change, and when are we only observing an association?

Marketing research courses often teach significance tests as procedures. A more useful approach teaches inferential reasoning: specify the comparison, state the estimand, protect randomisation, report uncertainty and connect statistical results to the size and value of the managerial effect. Experiments are especially important because digital marketing has made testing routine while also making false discovery easier to ignore.

What this concept covers

• Hypotheses, confidence intervals, p-values and effect sizes.
• Randomised experiments and A/B tests.
• Control groups, treatment integrity and pre-analysis choices.
• Multiple testing, peeking and false discovery.
• Causal versus associational claims and threats to validity.

Learning outcomes

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

• Formulate testable hypotheses linked to decisions.
• Interpret confidence intervals, significance and effect sizes.
• Design a simple randomised marketing experiment.
• Identify validity threats and multiple-testing risks.
• Translate experimental evidence into an economically meaningful recommendation.

Supports Course Outcomes 7 and 8.

How to teach it

• Use a result that is statistically significant but commercially trivial and another that is economically large but imprecise.
• Have students design an A/B test before showing them outcome data.
• Ask groups to state the stopping rule and primary metric in advance.
• Debrief what goes wrong when analysts continuously monitor and selectively report tests.

Case-style example

GreenVolt randomises 120,000 customers equally between a standard renewal email and a personalised message. Conversion rises from 3.20% to 3.55%, worth about £42 per retained customer, but the creative costs £18,000 to build and six secondary metrics were also checked. Students compute uplift, discuss confidence and estimate break-even value. One team may recommend rollout; another may demand replication because of multiple testing. Students discover that statistical significance is only one input to the decision.

Where students struggle

Students often read p < .05 as proof that an effect is important, or infer causality from regression without a defensible identification strategy. They also overlook pre-registration and stopping-rule discipline.

Reading and a quick check

Assign Feit and Berman (2019) on profit-maximising A/B tests and Berman and Van den Bulte (2022) on false discovery. Quick check: ask what additional information is needed before turning a statistically significant uplift into a rollout decision.

Applied teaching activity

Use a short case, research-design workshop, evidence critique or analyst memo so students leave the concept with a visible decision output rather than only lecture notes.

Where it leads

The next concept moves from single comparisons to multivariate structures, segmentation and prediction.

Central question: How do we reduce complex customer data into useful structures without mistaking an algorithmic output for a stable market truth?

Students do not need every multivariate technique at full mathematical depth, but they should understand what each method is for, what assumptions it makes and how to validate an interpretable result. Segmentation is especially important because it links marketing research to targeting and go-to-market decisions.

What this concept covers

• Correlation and regression as explanatory and predictive tools.
• Factor and dimension-reduction logic.
• Cluster analysis and segment profiling.
• Conjoint and preference trade-offs.
• Validation, holdouts, stability and managerial interpretability.

Learning outcomes

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

• Choose a multivariate method that fits the research question.
• Interpret coefficients, factors, clusters or utilities without overclaiming.
• Develop and profile a defensible segmentation.
• Test whether a model or segmentation is stable enough for decision use.
• Translate analytical outputs into customer and market actions.

Supports Course Outcomes 8 and 9.

How to teach it

• Show the same dataset clustered with different specifications and ask which differences matter managerially.
• Require students to name segments only after reviewing the variables that distinguish them.
• Use holdout validation or split-sample checks at an intuitive level.
• Ask groups to reject one technically "good" solution because it cannot support an action.

Case-style example

Orbit Travel surveys 2,400 customers on 12 attitudes, booking behaviour and willingness to pay. A factor analysis yields four dimensions and a three-cluster solution produces 34%, 41% and 25% segments. A four-cluster solution improves statistical fit but creates two commercially indistinguishable groups. Students profile both solutions, validate them on holdout data and recommend one for targeting. Students discover that the best segmentation balances evidence, stability, interpretability and actionability.

Where students struggle

Students can become impressed by model fit and forget that segments must be identifiable, substantial, reachable, stable enough and actionable. They also confuse predictive accuracy with causal explanation.

Reading and a quick check

Use Nunan, Birks and Malhotra, Chapters 22-26. Quick check: ask students why a statistically superior four-cluster solution might be worse for marketing than a simpler three-cluster solution.

Where a simulation fits

Go To Market Simulation: This concept prepares students for the Go To Market Simulation. After learning how segments are constructed and evaluated, students can use customer and market evidence to select one target segment and build a coherent positioning strategy rather than treating segmentation as an endpoint.

Where it leads

Modern research also draws on behavioural, digital and unstructured data that require different quality, privacy and interpretation checks.

Central question: How should researchers combine digital traces, behavioural data and text with declared survey responses without confusing what people did with why they did it?

Marketing researchers increasingly work with clickstreams, app events, CRM records, social content, reviews and text. These sources can reduce recall error and reveal behaviour at scale, but they create new problems of identity resolution, missing context, selection, platform dependence, privacy and opaque preprocessing. The course should teach integration rather than replace surveys with digital exhaust.

What this concept covers

• CRM, transaction, web, app and behavioural event data.
• Text, reviews, social content and qualitative-at-scale methods.
• Identity fragmentation, tracking limits and measurement error.
• Text coding, sentiment and human validation.
• Data linkage, governance, privacy and complementarity with primary research.

Learning outcomes

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

• Distinguish behavioural traces from attitudinal measures.
• Evaluate data provenance and coverage in digital datasets.
• Design a mixed evidence approach combining digital and primary research.
• Critique automated text or sentiment outputs.
• Identify privacy and governance risks in linked customer data.

Supports Course Outcomes 9 and 10.

How to teach it

• Compare what a clickstream can observe with what an interview can explain.
• Give students a sentiment model that misclassifies sarcasm and require a human validation plan.
• Ask students to map who is missing from a cookie- or app-based dataset.
• Use a data-linkage exercise where identifiers fragment across devices.

Case-style example

NovaBank has 50,000 anonymised app-event histories, 12,000 chat transcripts and a survey of 1,100 customers. The event data suggest frequent feature use, while interviews reveal customers often repeat actions because navigation is confusing. Students design an evidence-integration plan, sample 300 transcripts for validation and identify which data may be linked under the stated consent rules. One view trusts observed behaviour; another prioritises explanation. Students discover that behavioural volume does not eliminate construct and coverage problems.

Where students struggle

Students can assume digital data are objective because they are large and passively collected. They need to ask who generated the data, under what interface, with which identifiers and for what original purpose.

Reading and a quick check

Assign Tian, Hoban and Arora (2024) on limits of cookie-based measurement. Quick check: ask students to name one question behavioural data answer well, one they answer poorly and one they should not answer without additional consent.

Applied teaching activity

Use a short case, research-design workshop, evidence critique or analyst memo so students leave the concept with a visible decision output rather than only lecture notes.

Where it leads

Students can now integrate multiple evidence streams into market and competitive judgements that support actual marketing decisions.

Central question: How do research findings become a decision about which market to enter, which customers to prioritise and how to compete?

The value of marketing research is realised when evidence changes a decision. This concept integrates customer, market and competitor evidence into market attractiveness, segmentation, targeting, positioning and launch choices. Strategy frameworks can structure synthesis, but they should not replace evidence or encourage students to place the same fact into several boxes without asking what decision it informs.

What this concept covers

• Market attractiveness and customer opportunity.
• Competitor and category evidence.
• Segmentation, targeting and positioning as research-enabled decisions.
• PESTLE, Five Forces and SWOT as synthesis lenses.
• Insight-to-action logic, scenarios and recommendation criteria.

Learning outcomes

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

• Synthesize primary and secondary evidence into a market recommendation.
• Distinguish customer evidence, macro-environmental evidence and industry-structure evidence.
• Prioritise findings by materiality rather than novelty.
• Build an evidence-based STP and market-entry recommendation.
• Defend how the recommendation would change if one key assumption fails.

Supports Course Outcomes 9 and 10.

How to teach it

• Give different groups the same evidence but assign one PESTLE, Five Forces or SWOT lens, then compare what each framework makes visible and what it hides.
• Require a one-page "evidence -> insight -> implication -> decision" chain.
• Use a competitive launch problem so students must choose one target rather than recommend "serve everyone."
• Run the Go To Market Simulation only after the research analysis and segmentation work is complete.

Case-style example

LabFresh has four protein-bar segments: Performance (18%, high repeat, £2.60 price tolerance), Everyday Health (42%, medium repeat, £2.10), Convenience (25%, high channel accessibility, £1.90) and Budget (15%, £1.45). It can launch in only two of five US regions with one primary channel. Students combine segment economics, competitor density and qualitative needs, then recommend target, regions and positioning. Different routes can be defensible. Students discover that insight becomes valuable only when it forces trade-offs.

Where students struggle

Students often produce long lists of findings but no prioritisation, or use SWOT, PESTLE and Five Forces interchangeably. They need to match each framework to a distinct research question and then make a choice.

Reading and a quick check

Use Nunan, Birks and Malhotra, Chapters 28-29 alongside the course's market and competitor evidence. Quick check: ask students to convert one finding into an insight, an implication and a decision in four sentences.

Where a simulation fits

Go To Market Simulation: The Go To Market Simulation is the primary application here. Students use common customer and market evidence to segment the US protein-bar market, select one target segment and two launch regions, choose an access channel and build a differentiated positioning strategy. SWOT Analysis can be used as a secondary synthesis exercise; PESTLE Analysis and Porter's Five Forces fit when the learning goal is macro-environmental or industry-attractiveness judgement.

Where it leads

The final concept asks students to make the whole process ethically defensible, transparent about AI use and persuasive to decision makers.

Central question: What makes a marketing research recommendation trustworthy when data rights, AI assistance, uncertainty and stakeholder pressure all matter?

Ethics and communication should not be a final compliance lecture. Consent, privacy, fairness, disclosure and analytical transparency affect every stage of the research process. Generative AI adds useful tools for drafting, coding and exploration, but it also creates risks around fabricated evidence, confidentiality, synthetic respondents, reproducibility and the loss of an auditable human reasoning trail.

What this concept covers

• Consent, privacy, confidentiality and data minimisation.
• Researcher independence, incentives and vulnerable populations.
• AI-assisted instrument design, coding, summarisation and synthetic data risks.
• Open science, documentation, reproducibility and disclosure.
• Research reports, executive summaries, visual storytelling and oral defence.

Learning outcomes

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

• Apply ethical and privacy principles to a research design.
• Specify permitted and prohibited AI uses with disclosure requirements.
• Document an auditable research and analysis trail.
• Communicate uncertainty without weakening the decision recommendation.
• Defend a recommendation under methodological and managerial challenge.

Supports Course Outcome 10.

How to teach it

• Give groups an AI-assisted analysis and ask them to identify which steps can be verified and which cannot.
• Require a one-paragraph AI-use disclosure with every major assignment.
• Have students rewrite an overconfident executive summary so limitations are visible but the decision remains clear.
• Use a final oral defence that asks students what evidence would make them change their recommendation.

Case-style example

PulseHome asks a team to supplement 800 real panel responses with 2,000 AI-generated "synthetic consumers" because fieldwork is late. The synthetic data reduce apparent uncertainty and shift the preferred concept from B to A, while three interview transcripts contain personally identifiable information. Students decide which AI uses are permissible, what must be excluded, how data should be handled and how the final recommendation should be reported. Students discover that speed and apparent precision can conflict with validity, privacy and research integrity.

Where students struggle

Students may treat AI disclosure as a footnote and ethics as separate from validity. They need to see that undisclosed synthetic evidence, privacy breaches or irreproducible analysis directly weaken the credibility of the research conclusion.

Reading and a quick check

Assign Blanchard et al. (2025), Arora, Chakraborty and Nishimura (2025), and Deer et al. (2025). Quick check: ask students to write three permitted AI uses, three prohibited uses and the disclosure they would require in a module handbook.

Applied teaching activity

Use a short case, research-design workshop, evidence critique or analyst memo so students leave the concept with a visible decision output rather than only lecture notes.

Where it leads

This concept closes the lifecycle: students should now be able to define a decision problem, design and execute research, analyse evidence, recommend action and defend the integrity of the process.

10-week course structure

A 10-week version should compress adjacent stages rather than remove core research logic. Combine desk research with qualitative exploration, integrate measurement with piloting, and keep one applied go-to-market session plus a final ethics and communication session.

14-week course structure

A 14-week version creates space for separate qualitative analysis, deeper instrument development, more advanced multivariate interpretation and a fuller applied-simulation debrief. Use the extra time for evidence quality and student defence, not simply more techniques.

The Go To Market Simulation is a 1.5-2 hour, single-player competitive marketing simulation in which each student acts as the founder of a protein-bar company entering the US market. Students apply Segmentation, Targeting and Positioning to common customer and market evidence, then make connected launch choices. For Marketing Research, the value is not that the simulation replaces fieldwork. It tests whether students can turn evidence into a coherent decision and explain why that route follows from the data.

Why it fits here

By Session 11, students should already be able to evaluate segmentation evidence, compare market attractiveness, recognise data limitations and distinguish insight from recommendation. The simulation therefore becomes an application of completed research logic rather than an early guessing exercise.

What students do

• Begin with a defined protein-bar product, company mission, limited launch resources and a shared US market evidence base.
• Analyse customer motivations, buying behaviour, benefits sought and evidence quality to decide how the market should be divided.
• Select one primary target segment, two launch regions, a targeting approach, an access channel and a market-entry posture.
• Choose a positioning theme, frame of reference, point of difference, value proposition, price, pack format and promotional message.
• Enter a written reason for strategic choices so lecturers can compare the decision with the rationale.
• Review the final score, rank and strategy summary alongside cohort-level region, segment and pricing patterns, then identify where the route is coherent or internally inconsistent.

Recommended course placement

• At the end of Session 11 after segmentation and targeting have been taught.
• Within one two-hour applied class, with a separate pre-read and a short debrief.
• Before the final research or insight report so students can convert decisions into an evidence-to-action reflection.
• As an individual applied component in a larger group research course, because each participant submits a separate route to market.

Knowledge students should have before the simulation

• Segmentation bases and segment attractiveness.
• Target-market selection and the difference between size and fit.
• Positioning, value proposition and point of difference.
• How to read mixed qualitative and quantitative market evidence.
• How to distinguish an evidence-based rationale from a post-hoc explanation.

Delivery format

Suggested debrief questions

• Which evidence most affected your target choice, and which evidence did you discount?
• Where did your positioning choices reinforce or contradict the target customer evidence?
• Did a high-ranking route reflect strong research reasoning or only a favourable combination of choices?
• What additional research would most change your target, region, channel or price decision?
• Where did class patterns suggest genuine market consensus, and where might they reveal herding?
• Which part of the final strategy is most sensitive to an assumption rather than observed evidence?

The SWOT Analysis Simulation is a 3-4 hour, team-based strategy exercise in which Business Development and Risk Management teams assess the same international expansion opportunity from different priorities. Students analyse internal organisational evidence and external environmental conditions, classify and prioritise strengths, weaknesses, opportunities and threats, challenge competing interpretations and defend a recommendation. In a Marketing Research course, the learning value sits in evidence classification, materiality and synthesis rather than in memorising the four SWOT labels.

Why it fits here

By Session 10, students should be able to evaluate source quality, distinguish internal from external data and identify which findings are strong enough to change a decision. Role differences create a useful research debrief because the same evidence can produce different priorities without making either side automatically wrong.

What students do

• Review a role mandate and a shared international-expansion decision from either a Business Development or Risk Management perspective.
• Analyse internal evidence on resources, capabilities, financial position, operations and organisational constraints.
• Analyse external evidence on demand, competition, regulation, technology and wider environmental conditions.
• Classify evidence as strengths, weaknesses, opportunities or threats and prioritise the factors that are most material to the decision.
• Compare Business Development and Risk Management interpretations, challenge assumptions and reconcile where the same fact has been weighted differently.
• Defend an evidence-based strategic recommendation and use the result to identify what further research would reduce the most important uncertainty.

Recommended course placement

• After students have completed secondary research, qualitative/quantitative evidence interpretation and market synthesis.
• At the end of Session 10 as a bridge from analysis to the individual Go To Market application.
• Across two shorter classes where the first focuses on evidence classification and the second on challenge and recommendation.
• Before an executive research memo requiring students to distinguish evidence, interpretation and implication.

Knowledge students should have before the simulation

• The distinction between internal and external evidence.
• Source evaluation and materiality.
• How to move from a finding to an implication.
• Basic SWOT logic; a short introduction is enough for beginners.
• How role incentives and confirmation bias can affect interpretation.

Delivery format

Suggested debrief questions

• Which factor generated the largest disagreement between roles, and was the disagreement about evidence or weighting?
• Which item was misclassified first and why?
• What source would you commission next if one factor remained decision-critical but weakly evidenced?
• Did any team give more weight to evidence simply because it supported its role mandate?
• Which SWOT item should be removed because it is interesting but not material to the decision?
• How would the final recommendation change if the top-ranked opportunity or threat were wrong?

Grading with the Go To Market Simulation

The simulation gives each student a common evidence base and asks for an individual STP and launch strategy. Use the platform outcome as supporting evidence, then grade the quality of the research reasoning students can defend.

• Coherence between segmentation basis, chosen target, launch regions, channel and positioning.
• Use of customer and market evidence in the written rationale.
• Recognition of trade-offs, weak evidence and missing information.
• Ability to explain why a high or low score does not by itself establish the quality of the research reasoning.
• Individual post-simulation report or oral defence linking decisions to course methods.

Grading with the SWOT Analysis Simulation

Teams classify and prioritise internal and external evidence from different Business Development and Risk perspectives. The useful assessment signal is whether the recommendation follows from a transparent weighting of evidence.

• Correct distinction between internal and external factors.
• Evidence quality and the materiality of selected strengths, weaknesses, opportunities and threats.
• Prioritisation rather than exhaustive listing.
• Ability to identify role bias and reconcile competing interpretations.
• A short individual reflection on one factor the student would reclassify or reweight after the debrief.

Grading with the PESTLE Analysis Simulation

Growth Strategy and Risk Management teams score six macro-environmental dimensions, allocate weights that total 100% and bring their views into an Approval Committee. It works best as evidence for source evaluation and materiality.

• Quality of evidence used to justify each PESTLE score.
• Coherence of weights and whether the most material factors receive appropriate attention.
• Ability to distinguish a macro factor from an industry or company-specific factor.
• Contribution to the Approval Committee and consistency between analysis and enter, delay or do-not-enter recommendation.
• Individual note identifying one additional source that would most reduce uncertainty.

Grading with the Porter's Five Forces Simulation

Growth and Risk teams assess all five forces in the same premium-cafe market and defend a Go, Conditional Go or Reject decision. Grade the structural reasoning rather than the direction of the final recommendation.

• Application of all five forces as one connected profitability system.
• Written evidence supporting force scores and identification of interactions between forces.
• Distinction between market growth and structural industry attractiveness.
• Logical connection between the force analysis and the final recommendation.
• Individual ability to explain which assumption, if changed, would alter the industry-attractiveness judgement.

Platform evidence and individual attribution: The platform records submitted decisions, written rationales where the simulation provides them, scores or comparative outcomes across the cohort. That evidence supports your academic judgement; it does not replace it, and team-based platform outputs do not establish which individual student made which argument. Pair team evidence with an individual note, viva or reflection where an attributable mark is required.