4. Digital Research, Segmentation and Audience Intelligence

Introduction and Context

Digital and social media marketing is often described as data-driven, but the availability of data does not automatically produce insight. Platforms generate enormous quantities of behavioural, attitudinal and contextual information: searches, clicks, views, comments, shares, follows, purchases, location signals and device interactions. The strategic challenge is to determine which signals matter, what they mean, whose behaviour they represent and how confidently they can support a marketing decision.

Digital research provides the disciplined processes through which marketers investigate markets, audiences, cultures, needs and changing behaviour. Segmentation converts heterogeneous populations into meaningful groups. Audience intelligence combines research evidence, analytical interpretation and contextual understanding to explain not only who audiences are, but also what they value, how they make decisions, which communities influence them and how they respond to communication.

This chapter treats audience analysis as an interpretive and managerial activity rather than a purely technical exercise. It examines social listening, search data, platform insights, qualitative and quantitative research, segmentation, personas, communities and trend analysis. It also considers data quality, privacy, algorithmic bias and the limitations of digital evidence. These issues are central to strategic advertising because a poorly defined audience can cause wasted media investment, weak creative relevance and exclusionary or unethical targeting.

Learning Objectives

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

1. Explain the relationship between digital research, segmentation and audience intelligence.

2. Distinguish qualitative, quantitative, behavioural and contextual forms of audience evidence.

3. Use social listening, search data and platform analytics appropriately.

4. Develop useful segments and evidence-based personas.

5. Identify needs, communities, cultural meanings and emerging trends.

6. Evaluate data quality, representativeness, validity, reliability and bias.

7. Apply research findings to media, creative and strategic decisions.

8. Recognise ethical, legal and professional responsibilities in audience research.

From Audience Data to Audience Intelligence

Audience data are recorded observations, such as a search query, a video completion, a survey response or a purchase. Information is produced when data are organised and given context. Intelligence emerges when information is interpreted to support a decision. For example, knowing that a group has watched several videos about sustainable packaging is data. Concluding that the group will pay a premium for an environmentally responsible product may be an unjustified inference unless supported by additional evidence.

Audience intelligence therefore has three characteristics. First, it is decision-oriented: it addresses a marketing problem rather than collecting information for its own sake. Second, it is triangulated: it combines different sources because every source has blind spots. Third, it is critically evaluated: it recognises uncertainty, bias and the difference between correlation and causation.

A useful research chain is:

Business or communication question → research design → evidence collection → analysis → interpretation → strategic action → evaluation.

The chain is iterative. Campaign results may reveal a previously overlooked audience, while a social conversation may generate a new research question. Research should not be treated as a single preliminary stage that ends when a campaign begins.

Important Concepts and Terminology

A target population is the full group relevant to a research question. A sample is the subset from which evidence is collected. Reach refers to the number of distinct users exposed to content or advertising, while engagement refers to observable interactions such as comments, shares or saves. Intent is a stated or inferred likelihood of taking an action. Insight is an interpretation that explains a meaningful problem, motivation or opportunity and can guide action.

A first-party data source is collected directly by an organisation, such as its customer database, website analytics or loyalty programme. Second-party data are another organisation’s first-party data obtained through a partnership. Third-party data are collected or aggregated by external providers, often across multiple sources. The distinction matters because ownership, consent, accuracy and permissible use differ across sources.

Social listening is the systematic monitoring and analysis of digital conversations, mentions, themes, sentiment and networks. Social monitoring is generally narrower and focuses on responding to direct mentions or service issues. Search intent refers to the underlying purpose of a search, such as learning, comparing, locating or purchasing. A persona is a research-based representation of a strategically relevant audience type. A community is a group connected by shared identity, interest, practice or relationship, often with norms and forms of participation.

Research Questions and Evidence Design

Effective research begins with a precise question. “What do people think about our brand?” is broad and difficult to act upon. A stronger question might be: “What barriers prevent first-time users from considering the subscription, and which information sources reduce perceived risk?” The second question identifies an audience, a decision stage, a problem and a possible managerial use.

Research design should consider four dimensions. The first is the unit of analysis: an individual, household, post, search query, community, campaign or transaction. The second is the time frame: a short-term campaign period may reveal reactions, whereas longitudinal research is more suitable for habit or attitude change. The third is the source and method: platform data may measure behaviour, while interviews may explain motives. The fourth is the decision threshold: exploratory research may be sufficient for generating hypotheses, but major budget allocation may require stronger validation.

Qualitative Research

Qualitative research seeks depth, meaning and interpretation. It is useful when the organisation does not yet understand the language audiences use, when motivations are hidden, or when a category involves identity and culture. Methods include in-depth interviews, focus groups, online communities, digital ethnography, diary studies, observation of public conversations and analysis of user-generated content.

The value of qualitative research lies in its ability to reveal tensions. A person may describe a product as convenient but avoid it because using it publicly signals an undesirable identity. A focus group may show that a supposedly functional benefit is meaningful because it represents care for family members. Researchers commonly code transcripts or posts, identify recurring themes, compare cases and develop interpretive categories.

Qualitative findings should not be presented as numerical estimates of the entire market. A small number of interviews can reveal a mechanism or vocabulary, but cannot establish how common that mechanism is. The appropriate use of qualitative evidence is often hypothesis generation, message development, cultural interpretation and explanation of quantitative patterns.

Quantitative Research

Quantitative research measures incidence, distribution, relationships or change. Surveys can estimate attitudes, awareness, preference, satisfaction and stated purchase intention. Experiments can test the effect of different messages, formats or offers under controlled conditions. Digital analytics can describe behaviour at scale, including navigation paths, conversion events, retention and frequency.

A sound questionnaire uses clear constructs and avoids leading, double-barrelled or ambiguous questions. Measures should be aligned with the decision. If the objective is to understand trust, a single question such as “Do you trust the brand?” may be inadequate; several items may be required to capture credibility, transparency and reliability. Sampling procedures, response rates, missing data and question order should be documented.

Quantitative analysis may include descriptive statistics, cross-tabulation, correlation, regression, factor analysis or clustering. Advanced methods do not compensate for weak sampling or poorly defined variables. A statistically significant relationship may be practically trivial, and a large sample can make a small difference appear important. Analysts should report effect size, uncertainty and assumptions rather than relying only on a significance threshold.

Social Listening as Cultural and Market Research

Social listening gathers publicly available or appropriately licensed digital discourse to identify themes, issues, sentiment, language, influencers and networks. It can support brand health monitoring, competitor analysis, crisis detection, creative development, customer experience improvement and product innovation.

A robust listening project begins with a query design. The researcher selects brand names, product terms, competitor names, category language, common misspellings, hashtags and relevant cultural expressions. Exclusions are essential because a keyword may have several meanings. The query is then tested against a sample of results to identify irrelevant mentions, sarcasm, spam and automated activity.

Analysis may proceed through several layers. Volume indicates how much conversation exists, but not whether it is important. Share of voice compares a brand’s presence with competitors, while sentiment attempts to classify emotional direction. Topic analysis identifies issues associated with the brand. Qualitative reading is indispensable because automated sentiment tools often misinterpret irony, slang, code-switching and context. Network analysis can identify clusters and bridges between communities, but visible popularity should not be confused with influence over actual behaviour.

Social listening is strongest when it connects discourse with organisational action. Repeated complaints about confusing delivery information may lead to service redesign, not merely a change in advertising copy. Similarly, a growing vocabulary around repair and reuse may indicate an opportunity for content or product innovation, but it requires validation beyond conversation volume.

Search Data and the Study of Intent

Search behaviour offers a valuable view of active information needs. Unlike many forms of passive exposure data, a search query often records a question or problem that a person is trying to solve. Search data can reveal category vocabulary, seasonal demand, comparison behaviour, geographic variation and stages of the decision journey.

Search terms may be interpreted through intent categories. Informational queries seek knowledge; navigational queries seek a particular site or destination; commercial investigation queries compare options; and transactional queries indicate readiness to act. These categories are not rigid, and the same person may move among them. Search volume is also not equivalent to market demand: a term may be popular because of news, curiosity or controversy rather than purchase interest.

A researcher should examine related queries, trends over time, landing-page behaviour and conversion quality. High traffic with low engagement may indicate a mismatch between search promise and page content. A lower-volume query may be strategically valuable if it reflects strong intent and produces qualified customers. Search data should therefore be interpreted alongside customer interviews, web analytics and sales outcomes.

Platform Insights and Behavioural Evidence

Platform insights provide information about audience composition, content performance and media delivery. Depending on the platform and permissions, marketers may examine age ranges, broad locations, device use, content interactions, watch time, retention, click-through rates, frequency and conversion events. These data help answer operational questions: Which creative format holds attention? Which audience responds to a message? At what point do viewers leave? Which placements produce efficient outcomes?

Platform analytics are not neutral portraits of society. They are shaped by platform definitions, tracking systems, user declarations, algorithmic distribution and the organisation’s own tagging choices. “Engagement” may reward controversy, while a quiet exposure may influence later behaviour without producing a visible interaction. Reported demographic categories can be incomplete or inferred. Consequently, platform insights are best treated as evidence about platform activity, not as a complete census of the audience.

Segmentation: Moving Beyond Demographics

Segmentation divides a broad market into groups that are sufficiently similar within themselves and sufficiently different from one another to justify distinct strategic treatment. Traditional bases include geographic, demographic, socioeconomic and psychographic characteristics. Digital environments make it possible to add behavioural, needs-based, journey-based, contextual and value-based segmentation.

Demographic variables can be useful for media planning and market description, but they rarely explain behaviour on their own. Two people of the same age and income may have different motivations, media habits and cultural identities. Behavioural segmentation may distinguish frequent users from occasional users, while needs-based segmentation may distinguish risk reducers from convenience seekers. Journey segmentation may separate discovery, evaluation, conversion, use and advocacy.

A segment should be evaluated against practical criteria. It should be identifiable through legitimate evidence, substantial enough to matter, reachable through available channels, meaningfully different in needs or response, and actionable through an appropriate proposition. A segment that is easy to name but impossible to reach or serve differently is not strategically useful.

A Practical Segmentation Process

The process normally begins by defining the strategic problem and assembling relevant data. Researchers then explore variables, remove unreliable measures and identify patterns. Statistical clustering can assist in grouping respondents, but the resulting clusters require managerial interpretation. A segment profile should describe its needs, barriers, behaviours, triggers, preferred information sources, relationship with the category and likely response to value propositions.

Segments should be tested for stability and usefulness. If a small change in the sample produces completely different groups, the segmentation may be unstable. If all segments respond similarly to the same message, differentiation may not justify additional complexity. Segmentation should be revisited when the market, product, technology or social context changes.

Personas, Needs and Communities

Personas translate research findings into a memorable representation for creative, media and product teams. A credible persona is not a fictional biography filled with invented details. It should state the evidence base, identify the person’s context and show the tension that affects action. A useful persona may include a goal, unmet need, barrier, decision trigger, trusted sources, behavioural evidence, representative language and implications for communication.

Needs can be functional, emotional, social or symbolic. Functional needs concern what a product does; emotional needs concern how the person wishes to feel; social needs concern relationships and recognition; symbolic needs concern identity and cultural meaning. The same product benefit may satisfy several needs. For example, a time-saving service may support efficiency, reduce anxiety and allow a parent to feel more present with family.

Communities require particular care. A community is not simply an audience selected by demographic criteria. It has shared meanings, practices, norms and sometimes internal disagreement. Brand participation should begin with observation and respect. Marketers should understand who speaks, who is marginalised, what forms of contribution are valued and whether commercial involvement is welcome. Community insight can inform language and partnership, but extracting cultural symbols without accountability can damage trust.

Trend Research and the Difference Between Signal and Noise

A trend is a sustained or developing pattern of change, whereas a spike may be temporary attention. Trend research combines time-series data, search behaviour, social discourse, cultural observation, category performance and expert interpretation. Researchers should ask whether a pattern is growing, who is driving it, what conditions sustain it and whether it changes behaviour or merely vocabulary.

Trend analysis is vulnerable to recency bias and platform visibility bias. A highly visible online discussion may not represent wider public adoption. Conversely, a slow change in household practice may receive little social attention. Useful trend reports distinguish evidence from interpretation, identify confidence levels and present alternative explanations. Scenario planning can then explore how a trend might affect consumers, competitors, regulation and brand choices.

Data Quality, Bias and Research Limitations

The quality of audience intelligence depends on more than data volume. Validity concerns whether a method measures what it claims to measure. Reliability concerns consistency. Representativeness concerns whether evidence reflects the relevant population. Completeness concerns missing cases and fields. Timeliness concerns whether the data remain appropriate for the decision. Provenance concerns where data came from and how they were processed.

Digital research can contain several forms of bias:

• Coverage bias: people without access to, or participation in, a platform are absent.
• Self-selection bias: highly satisfied or dissatisfied users may be more likely to post or respond.
• Measurement bias: platform metrics may use definitions that differ from the researcher’s construct.
• Algorithmic bias: ranking and recommendation systems make some voices more visible than others.
• Demographic bias: inferred or incomplete profiles may misclassify people.
• Survivorship bias: analysis of retained users ignores those who abandoned a service.
• Confirmation bias: analysts notice evidence that supports an existing belief.
• Response bias: respondents may give socially desirable answers.

Triangulation reduces, but does not eliminate, these risks. Researchers should compare data sources, inspect unusual cases, document exclusions, test alternative definitions and report limitations. A dashboard that presents precise numbers without methodological notes can create false confidence.

Correlation deserves special caution. If exposed users convert at a higher rate than unexposed users, exposure may have contributed, but the groups may also differ in prior interest, income or intent. Randomised experiments, credible control groups, geo-tests or carefully designed quasi-experiments provide stronger evidence of causation than descriptive platform comparisons.

Ethical, Legal and Governance Considerations

Audience research operates within a relationship of trust. Personal data should be collected and used for specified, legitimate purposes, with appropriate consent or another lawful basis. Organisations should follow applicable privacy and data-protection requirements, provide transparency, minimise collection, protect data and respect deletion or access rights where applicable. Public availability does not automatically make every use ethically acceptable.

Sensitive attributes require heightened care. Targeting or inference based on health, ethnicity, political belief, religion, sexuality or financial vulnerability can create discrimination and personal harm. Even apparently neutral variables may function as proxies. Researchers should conduct privacy and impact assessments, restrict access, use aggregation where possible and avoid targeting that exploits vulnerability.

Ethical practice also includes honest reporting. Researchers must not remove inconvenient findings, manufacture personas, manipulate sentiment classifications or present estimates as facts. When using online communities, researchers should consider whether participants reasonably expect observation, whether quotations could identify them and whether consent is required. Governance should assign responsibility for data stewardship, model review, retention and escalation of harmful findings.

Technology, Automation and Artificial Intelligence

Digital research platforms can automate collection, classification, clustering, transcription, topic modelling, dashboarding and predictive scoring. Machine-learning systems can identify patterns in large datasets that would be difficult to inspect manually. Generative AI can assist with coding, summarising interviews or proposing hypotheses, but it should not replace source verification or researcher judgement.

Automated systems may reproduce training-data bias, misread context, infer sensitive characteristics or generate plausible but unsupported interpretations. Human review is especially important for sarcasm, minority dialects, multilingual content, crisis signals and high-impact decisions. A professional workflow records the prompt or model version where relevant, preserves source material, checks representative samples and distinguishes machine-generated hypotheses from validated findings.

An Illustrative Case: Repositioning a Plant-Based Beverage

Background

The following fictional case concerns “Natura,” a plant-based beverage brand experiencing strong online awareness but weak repeat purchase. The brand team initially believed that younger urban consumers formed one homogeneous target audience.

Situation and Problem

Sales data showed trial concentrated in large cities, while repeat purchase varied considerably. Social listening revealed positive conversation about environmental benefits but frequent comments about taste, price and uncertainty about how the product should be used. Search data showed queries related to coffee recipes, protein and lactose intolerance, while platform analytics indicated that recipe videos generated more saves than image-based sustainability posts.

Analysis

The team commissioned twelve online interviews with recent purchasers, a survey of 1,200 category users and an analysis of three months of public category conversation. The interviews revealed three distinct tensions: some users wanted a reliable coffee ingredient, some sought a convenient nutrition solution and some wanted to reduce dairy consumption without sacrificing taste. The survey indicated that these motivations were not distributed equally across age or income groups. Social listening also showed that environmental claims attracted attention but rarely answered the practical question of product use.

The evidence had limitations. The survey used an online panel and therefore might underrepresent non-digital consumers. Social conversation overrepresented enthusiasts and critics. Platform data described content interactions rather than total brand influence. These limitations were recorded in the research report.

Approach and Strategy

The team developed three provisional segments: “routine replacers,” seeking a dependable everyday substitute; “performance planners,” seeking nutritional convenience; and “curious experimenters,” motivated by recipes and novelty. Personas were created only after linking each segment to survey patterns, interview quotations and observed behaviour. The brand tested separate creative propositions through a controlled campaign design. Routine replacers saw demonstrations of coffee performance, performance planners saw portable usage occasions, and experimenters saw recipe collaborations.

The team also changed search landing pages so that practical queries led to relevant recipes and product explanations. Community partnerships were selected for demonstrated credibility rather than follower count alone. Measurement included qualified site visits, product trial, repeat purchase and brand association with taste and usability.

Outcome

In this illustrative case, the brand observed that recipe-led communication produced stronger saving and return visits than environmental awareness content, while practical demonstrations improved product consideration among routine replacers. The result did not prove that one segment was permanently more valuable; it showed that different needs required different evidence and propositions. The brand continued testing because the research period could not establish long-term behaviour change.

Learning Points

The case demonstrates five principles: demographic labels can conceal different needs; behavioural data require interpretation; qualitative research explains why patterns occur; segments should lead to differentiated action; and research limitations must accompany findings. It also shows that audience intelligence is a cross-functional activity linking insight, creative, media, product and measurement teams.

Professional Practice and Best Practices

A professional researcher maintains a clear research brief, records definitions and decisions, and distinguishes observed behaviour from inferred motivation. The team should combine at least two materially different evidence sources when making important decisions. It should predefine success measures, test data quality, involve cultural and privacy review where necessary and communicate uncertainty to managers.

A practical audience-intelligence report normally includes the business question, methods, population and sample, period of analysis, definitions, major findings, evidence excerpts, segment implications, recommended actions, confidence assessment and limitations. Visualisations should show denominators and time periods. Personas should cite their evidence base. Dashboards should not encourage teams to optimise only what is easiest to count.

The most useful output is not the largest database or the most elaborate segment model. It is a defensible explanation of audience reality that improves a decision and can be revised when new evidence appears.

Current Developments and Emerging Directions

Privacy changes are encouraging greater reliance on first-party relationships, consent-based data and aggregated measurement. Media planners increasingly work with clean rooms, contextual signals, modeled conversions and incrementality testing rather than assuming person-level tracking is always available or desirable. Social commerce is bringing behavioural, content and transaction data closer together, while creator and community research is expanding beyond follower counts toward trust, participation and audience fit.

AI-assisted research will increase the speed of analysis, but the strategic premium will shift toward problem definition, ethical judgement, cultural literacy, validation and causal reasoning. Synthetic audiences and predictive models may help explore scenarios, yet they remain models rather than substitutes for lived human evidence. Future-ready practitioners will be able to combine technical fluency with scepticism about what data cannot show.

Chapter Summary

Digital research converts observations into evidence for marketing decisions, while audience intelligence interprets that evidence in context. Social listening reveals public discourse and cultural meaning; search data indicate active information needs; platform insights describe interactions and delivery; qualitative methods explain motivations; and quantitative methods estimate patterns and relationships. Effective segmentation moves beyond demographics to needs, behaviours, journeys and values. Personas should be evidence-based, and communities should be approached as social systems rather than convenient targeting labels.

Research quality depends on validity, reliability, representativeness, provenance, timeliness and transparent limitations. Digital evidence is affected by coverage, self-selection, algorithmic, measurement and confirmation biases. Triangulation and experimentation strengthen decisions but do not remove uncertainty. Ethical practice requires privacy protection, responsible use of sensitive data, fairness, transparency and human oversight of automated systems. The professional objective is not simply to know more about audiences, but to make better, more accountable strategic choices.

Key Terms

Audience intelligence: Interpreted audience evidence used to guide strategic decisions.

Behavioural data: Recorded evidence of actions such as clicks, views, searches or purchases.

Community: A group connected by shared identity, interest, practice or relationships.

Correlation: A statistical association between variables that does not by itself establish causation.

Data provenance: The documented origin and processing history of data.

First-party data: Data collected directly by an organisation from its own interactions.

Insight: An evidence-based interpretation of a meaningful problem, motivation or opportunity.

Intent: The purpose or likely direction underlying a behaviour, especially a search or decision.

Persona: A research-based representation of a strategically relevant audience type.

Qualitative research: Research focused on meanings, experiences, motivations and context.

Quantitative research: Research that measures distributions, relationships, incidence or change numerically.

Representativeness: The degree to which evidence reflects the population of interest.

Sample: A subset of a population selected for research.

Segmentation: The division of a market into groups with strategically meaningful similarities and differences.

Search intent: The underlying purpose of a user’s search query.

Social listening: Systematic analysis of digital conversations, themes, sentiment and networks.

Triangulation: Comparing multiple methods or data sources to strengthen interpretation.

Validity: The extent to which a measure captures the concept it is intended to measure.

Reliability: The consistency of a measure or research procedure.

Trend: A sustained or developing pattern of change rather than a temporary spike.

Review Questions

Short Answer Questions

1. Distinguish data, information and audience intelligence.

2. What is the difference between social monitoring and social listening?

3. Why should demographic variables not be used as the sole basis for segmentation?

4. Define validity and reliability in digital research.

5. What makes a persona evidence-based?

Long Answer Questions

1. Explain how qualitative and quantitative research can be combined in an audience-intelligence project.

2. Discuss the uses and limitations of social listening for advertising strategy.

3. Describe a practical process for developing and evaluating market segments.

4. Explain how search data can reveal consumer intent while still producing misleading conclusions.

5. Discuss the ethical and governance issues involved in using personal and sensitive audience data.

Analytical / Application Questions

1. A campaign receives high engagement but produces few qualified leads. Design a research plan using platform insights, interviews and search data to diagnose the problem.

2. A brand’s social conversation suddenly doubles after a controversial news story. Explain how you would determine whether this is a meaningful trend, a temporary spike or a measurement artefact.

3. Develop two possible needs-based segments for a digital financial service and specify the evidence, messages, channels and ethical safeguards required for each.

22. Multiple Choice Questions with Answers

Question 1

Which statement best distinguishes audience intelligence from audience data?

A. Audience intelligence is always collected through surveys

B. Audience intelligence interprets evidence to support a decision

C. Audience intelligence consists only of demographic information

D. Audience intelligence is limited to platform engagement metrics

Correct Answer: B. Audience intelligence interprets evidence to support a decision

Question 2

A researcher finds that users who saw an advertisement converted more often than users who did not. What is the most appropriate conclusion from this observation alone?

A. The advertisement definitely caused all additional conversions

B. The exposed users may have differed from the unexposed users in other ways

C. The campaign should immediately be scaled without further testing

D. Conversion data are irrelevant to audience research

Correct Answer: B. The exposed users may have differed from the unexposed users in other ways

Question 3

Which research method is most appropriate for discovering the language consumers use to describe an unfamiliar category?

A. Exploratory interviews

B. A frequency report based only on impressions

C. A conversion dashboard without search terms

D. A demographic media schedule

Correct Answer: A. Exploratory interviews

Question 4

Why can high social-conversation volume be misleading?

A. Volume always excludes competitor mentions

B. Volume proves positive sentiment

C. Volume may reflect controversy, spam or a small highly active group

D. Volume measures only offline purchases

Correct Answer: C. Volume may reflect controversy, spam or a small highly active group

Question 5

Which characteristic makes a market segment strategically useful?

A. It has an elaborate name and many descriptive details

B. It is actionable and meaningfully different in needs or response

C. It is defined only by age

D. It changes completely whenever a new post appears

Correct Answer: B. It is actionable and meaningfully different in needs or response

Question 6

A search term receives modest traffic but produces highly qualified visits and purchases. How should the marketer interpret it?

A. It has no value because traffic is low

B. It may indicate valuable high-intent demand

C. It must be a branding term

D. It should be removed from analysis

Correct Answer: B. It may indicate valuable high-intent demand

Question 7

Which is an example of coverage bias?

A. Respondents answering a sensitive question inaccurately

B. A sentiment tool misreading sarcasm

C. People who do not use a platform being absent from its audience data

D. An analyst preferring a familiar hypothesis

Correct Answer: C. People who do not use a platform being absent from its audience data

Question 8

What is the most responsible use of an AI system that classifies online conversation?

A. Accept every classification without inspection

B. Use classifications as hypotheses and validate them with human review

C. Infer sensitive characteristics for targeting automatically

D. Delete all source material after generating a summary

Correct Answer: B. Use classifications as hypotheses and validate them with human review

Question 9

Which statement about a research-based persona is correct?

A. It should contain invented biographical details to feel realistic

B. It represents every member of the market identically

C. It links audience context, needs and barriers to evidence

D. It replaces the need for segment evaluation

Correct Answer: C. It links audience context, needs and barriers to evidence

Question 10

A brand observes a sudden hashtag spike lasting two days. What should the researcher do first?

A. Declare a permanent cultural trend

B. Ignore it because online data are never useful

C. Examine its source, duration, composition and behavioural significance

D. Target everyone who used the hashtag

Correct Answer: C. Examine its source, duration, composition and behavioural significance