Market Research Glossary: Key Terms and Definitions
Market research has developed a precise vocabulary over decades of applied practice, academic refinement, and methodological debate. The terms below represent the working language of the discipline — used by brand strategists, consumer insight teams, product developers, and policy researchers. Definitions are given in their professional sense, not their colloquial approximations.
A/B Testing
A controlled experiment in which two versions of a stimulus — an advertisement, a product page, a pricing structure, a survey question — are exposed to separate audience segments simultaneously. The goal is to isolate the effect of a single variable. A/B testing is a form of causal inference by design; its validity depends on random assignment, adequate sample size, and the absence of confounding between groups.
Aided Recall
A measurement technique in which respondents are given a cue — typically a brand name, logo, or category prompt — before being asked whether they recognize or remember it. Aided recall measures recognition rather than spontaneous memory. It consistently produces higher response rates than unaided recall and is used when researchers need to assess depth of brand penetration rather than top-of-mind salience.
Attitude Scale
A structured instrument used to quantify subjective opinions, feelings, or evaluations. Common formats include the Likert scale (degree of agreement), the semantic differential (bipolar adjective pairs), and the Thurstone scale (pre-calibrated statements). Attitude scales convert qualitative psychological states into ordinal or interval data suitable for statistical analysis.
Brand Equity
The aggregate value that a brand name adds to a product or service beyond its functional attributes. Brand equity is not a balance sheet figure in most accounting frameworks, but it is operationalized in market research through four components: brand awareness, perceived quality, brand associations, and brand loyalty. David Aaker’s model remains the dominant framework for measuring it.
Category Development Index (CDI)
A ratio comparing a product category’s sales performance in a specific geographic or demographic segment to that segment’s share of the total population. A CDI above 100 indicates that the category over-indexes in that segment. CDI is used alongside the Brand Development Index (BDI) to prioritize media spend and identify underpenetrated markets.
Cluster Analysis
A multivariate statistical method that groups respondents, brands, or products into segments based on the similarity of their attribute profiles. Unlike regression, cluster analysis is unsupervised — it does not presuppose a dependent variable. It is widely used in consumer segmentation studies to derive typologies from survey data, behavioral logs, or transactional records.
Concept Test
A research method used to evaluate consumer response to a new product or service idea before development investment is committed. Respondents are exposed to a concept statement — a written, visual, or prototype description of the proposed offering — and assessed on purchase intent, perceived uniqueness, and fit with stated needs. Monadic, sequential monadic, and paired comparison designs are the standard formats.
Conjoint Analysis
A technique for decomposing consumer preferences into the relative utilities of individual product attributes. Respondents evaluate product profiles composed of varying attribute combinations; statistical modeling then derives the implicit trade-offs they are willing to make. Conjoint analysis is used to simulate market share under different product configurations and to identify the attribute levels that maximize willingness to pay.
Consumer Panel
A standing sample of individuals or households recruited to provide repeated observations over time. Panels enable longitudinal tracking of purchase behavior, brand switching, media consumption, and attitudinal change. Panel data is used to calculate household penetration, purchase frequency, and category share with greater precision than cross-sectional surveys.
Crosstabulation
A table that presents the frequency distribution of one variable broken down by one or more other variables. Crosstabs are the standard output format for survey data analysis and are used to identify differences in response patterns across demographic, psychographic, or behavioral subgroups. Chi-square tests are applied to assess whether observed differences are statistically significant.
Demand Curve Estimation
The process of determining the relationship between price and quantity demanded for a product or category, using survey-based methods (Gabor-Granger, Van Westendorp), observational sales data, or experimental variation. Demand curve estimation is foundational to pricing research and to calculating price elasticity.
Diary Study
A qualitative or quantitative method in which participants record their behavior, experiences, or thoughts in structured logs over an extended period — typically days or weeks. Diary studies capture in-context data that retrospective interviews miss and are particularly valuable for studying media habits, shopping behavior, health routines, and product usage in naturalistic settings.
Ethnography
An observational method derived from cultural anthropology in which a researcher embeds in the environment of the subjects under study to document behavior, norms, and social dynamics as they occur. In market research, ethnographic methods range from in-home observation sessions to extended field studies. Ethnography surfaces the gap between what consumers say they do and what they actually do.
Focus Group
A moderated discussion involving six to ten participants recruited on the basis of shared characteristics. Focus groups generate qualitative data — language, emotional responses, group dynamics, and associative thinking — that surveys cannot capture. Their limitation is dual: small samples preclude quantitative generalization, and group dynamics can suppress minority opinions. Focus groups are best used to generate hypotheses, develop survey instruments, and explore unanticipated dimensions of a topic.
Gabor-Granger Method
A pricing research technique in which respondents are asked sequentially whether they would purchase a product at a given price, with the price varied across respondents or across rounds. The method produces a demand curve and an optimal price point that maximizes revenue. It assumes that demand decreases monotonically with price, which limits its applicability in markets where price signals quality.
Halo Effect
A cognitive bias in which a respondent’s overall impression of a brand, product, or person influences their evaluation of specific attributes. A positive halo leads to inflated ratings on individual dimensions; a negative halo suppresses them. The halo effect is a systematic measurement error in attitude research and is addressed through question sequencing design and analytical controls.
Incidence Rate
The proportion of the general population that meets the screening criteria for a given study. A study targeting frequent premium whisky buyers among adults aged 35–54 will have a lower incidence rate than one targeting all adult beverage consumers. Incidence rate directly determines fieldwork cost and feasibility; studies with incidence rates below 5% require either specialized panels or significantly higher per-respondent incentives.
Key Driver Analysis
A technique used to identify which product or brand attributes have the greatest statistical influence on an outcome of interest, typically overall satisfaction or purchase intent. Methods include multiple regression, relative weights analysis, and Shapley value decomposition. Key driver analysis prioritizes investment by distinguishing attributes that matter from those that merely correlate with positive outcomes.
Laddering
A qualitative interviewing technique that probes progressively from product attributes to personal values. The interviewer asks “why is that important to you?” in repeated iterations following an initial response, tracing a chain from concrete features to functional consequences to abstract motivations. Laddering maps the means-end structure of consumer decision-making and is used to develop brand positioning and communication strategy.
Latent Class Analysis (LCA)
A model-based segmentation method that identifies unobserved subgroups within a population based on patterns of response to categorical variables. Unlike cluster analysis, LCA provides probabilistic class membership and fit statistics that allow comparison across model specifications. It is used in segmentation research when the underlying population structure is not directly observable.
Market Sizing
The process of estimating the total revenue or unit volume available to a product or service in a defined market. Top-down sizing begins with macro-level data and applies penetration assumptions; bottom-up sizing aggregates unit-level demand estimates. Market sizing is an input to investment decisions, category entry assessments, and demand planning.
MaxDiff (Best-Worst Scaling)
A scaling technique in which respondents choose the best and worst item from subsets of a larger attribute list. By analyzing the pattern of choices across multiple subsets, researchers derive interval-level preference scores for all items. MaxDiff avoids the response biases associated with rating scales — particularly acquiescence bias and scale compression — and produces data suitable for priority ranking and trade-off analysis.
Net Promoter Score (NPS)
A single-question metric that asks respondents how likely they are to recommend a brand or product to others, on an eleven-point scale from zero to ten. Respondents scoring nine or ten are classified as Promoters; those scoring zero through six as Detractors. NPS is calculated as the percentage of Promoters minus the percentage of Detractors. It is widely used as a customer loyalty indicator, though its predictive validity and theoretical foundations remain contested in the academic literature.
Omnibus Survey
A regularly conducted survey shared across multiple clients or research questions, with each client purchasing space for a limited number of questions. Omnibus surveys provide cost-efficient access to nationally representative samples and are used for tracking studies, quick-turnaround fact-finding, and situations where sample size requirements exceed what a custom study can justify.
Panel Fatigue
The degradation of data quality that occurs when panel members respond to surveys frequently over time. Fatigued respondents straightline through scales, select socially desirable answers, or abandon questionnaires early. Panel fatigue is managed through recruitment rotation, incentive structures, survey length controls, and data quality screening metrics applied at the individual respondent level.
Perceptual Mapping
A graphical representation of how brands or products are positioned in the minds of consumers relative to each other and relative to defined attribute dimensions. Perceptual maps are derived from correspondence analysis, multidimensional scaling, or factor-analytic methods applied to attribute rating data. They are used to identify competitive white space, diagnose positioning drift, and evaluate repositioning strategies.
Primary Research
Data collected directly by or for a specific study, from original sources. Surveys, interviews, focus groups, observations, and experiments are all forms of primary research. Primary research is distinguished from secondary research by its purpose-built design; its advantage is specificity, its disadvantage is cost and time.
Projective Technique
An indirect elicitation method that asks respondents to interpret or complete an ambiguous stimulus — a sentence, a picture, a story — as a proxy for their own attitudes or motivations. Projective techniques are used when direct questioning is unlikely to surface honest or self-aware responses, such as in studies of embarrassing purchases, socially sensitive behaviors, or unconscious brand associations.
Psychographics
Variables used to segment populations based on personality traits, values, attitudes, interests, and lifestyle patterns. Psychographic segmentation goes beyond demographic classification to describe the psychological characteristics that shape purchasing behavior and media consumption. VALS (Values and Lifestyles) and Experian Mosaic are among the commercially available psychographic frameworks applied in market research.
Qualitative Research
Research methods that generate non-numerical data — language, imagery, behavior, and narrative — and interpret it through analytic rather than statistical means. Qualitative research explores the why and how behind behavior; it is generative, contextual, and hypothesis-producing rather than hypothesis-testing. It cannot produce statistically valid frequency estimates.
Quantitative Research
Research methods that produce numerical data capable of statistical analysis and generalization to a defined population. Quantitative research tests hypotheses, measures incidence, and tracks change over time at scale. Its validity depends on probability-based sampling, controlled measurement conditions, and adequate sample sizes.
Regression Analysis
A statistical method that models the relationship between a dependent variable and one or more independent variables. In market research, regression is applied to predict outcomes (purchase intent, satisfaction), decompose drivers, and control for confounding variables. Ordinary least squares, logistic, and ordinal regression are the most commonly used variants depending on the scale of the dependent variable.
Respondent Bias
Any systematic distortion in survey responses attributable to characteristics of the respondent rather than the true construct being measured. Forms include acquiescence bias (tendency to agree regardless of content), social desirability bias (tendency to give normatively acceptable answers), extreme response bias (tendency to select endpoints of scales), and demand characteristics (tendency to respond in ways consistent with perceived researcher expectations).
Sample Frame
The operational list or population from which a study’s sample is drawn. The sample frame defines who can be selected; its coverage relative to the target population determines whether sampling error or coverage error is the greater threat to validity. A telephone sample frame, for example, excludes households without phones and introduces systematic coverage bias if that exclusion is not random.
Secondary Research
The analysis of data originally collected for purposes other than the current study. Syndicated data, government statistics, academic publications, industry reports, and prior proprietary research are all secondary sources. Secondary research is faster and cheaper than primary research and is used to establish market context, benchmark performance, and identify gaps that primary research must fill.
Segmentation
The process of dividing a market into distinct subgroups of consumers who share common characteristics relevant to product choice, communication response, or value delivery. Segmentation variables include demographics, geographics, psychographics, and behavioral data. Actionable segmentation is defined by four criteria: segments must be measurable, substantial, accessible, and differentially responsive to marketing stimuli.
Share of Wallet
The proportion of a consumer’s total category spending that goes to a specific brand or supplier. Share of wallet is a behavioral measure of loyalty and is used alongside satisfaction metrics to identify customers who are satisfied but nonetheless diversified across competitors. Increasing share of wallet from existing customers is typically more cost-efficient than acquiring new ones.
Simulated Test Market (STM)
A controlled research method that models a product’s likely sales performance under defined marketing conditions without a full product launch. Respondents are exposed to advertising, given an opportunity to purchase in a simulated retail environment, and then resurveyed on repurchase intent and usage. STM models — including BASES and Litmus — use the behavioral data to project trial and repeat purchase rates.
Stated Preference
A research approach in which consumer preferences are inferred from responses to hypothetical choice scenarios rather than from observed behavior. Conjoint analysis, discrete choice experiments, and willingness-to-pay surveys are stated preference methods. They are used when revealed preference data is unavailable — typically in new product development, pricing research, and policy analysis.
Syndicated Data
Standardized research data collected continuously by a third-party supplier and sold to multiple clients. Nielsen retail scanner data, IRI consumer panel data, and GfK media measurement are examples. Syndicated data provides industry-wide benchmarks, competitive context, and longitudinal trend information that proprietary studies cannot replicate at equivalent cost.
Target Population
The total group of individuals or entities to which the conclusions of a research study are intended to apply. The target population is defined by inclusion and exclusion criteria — typically demographic, behavioral, or geographic — established before sampling begins. The sample is drawn from the target population; inferences from the sample are valid only to the extent that the sample is representative of it.
Tracking Study
A longitudinal research design in which the same set of measures is administered to comparable samples at regular intervals — monthly, quarterly, or annually. Tracking studies monitor changes in brand awareness, image attributes, advertising recall, and purchase behavior over time and across competitive events. Consistency of methodology is the primary constraint: changes in questionnaire wording, sample composition, or fieldwork procedure introduce measurement artifacts that confound trend analysis.
Unaided Recall
A measurement technique in which respondents are asked what brands or products they can remember in a category without being provided any cues. Unaided recall measures top-of-mind awareness and indicates the depth of a brand’s presence in long-term memory. It consistently produces lower response rates than aided recall and is regarded as a more demanding and more meaningful test of brand salience.
Validity
The degree to which a research instrument measures what it claims to measure. Internal validity refers to the accuracy of causal inference within a study; external validity refers to the generalizability of findings to other populations and contexts. Construct validity concerns whether the operationalization of a concept correctly captures the underlying theoretical variable. Market research instruments are assessed on all three dimensions.
Van Westendorp Price Sensitivity Meter
A pricing research technique that uses four questions to identify acceptable price ranges for a product. Respondents identify the price at which a product would be considered too cheap to be trustworthy, a bargain, expensive but worth considering, and too expensive to purchase. The intersections of these cumulative frequency distributions define the range of acceptable pricing and the optimal price point. Unlike Gabor-Granger, Van Westendorp does not assume monotonic demand.
Willingness to Pay (WTP)
The maximum price a consumer would pay for a specific product or service. WTP can be elicited through direct survey questions, incentive-compatible auction mechanisms, or modeling from conjoint data. It is a key input to pricing strategy, value proposition development, and new product forecasting. Direct WTP questions are subject to hypothetical bias — respondents consistently overstate WTP relative to actual purchase behavior.