We at FGI want our website to serve as a knowledge-base as much as a venue to inform you of our offerings. We put together this market research glossary, which is constantly evolving and adding new terms, as a way to increase your understanding and showcase the depth of our research expertise.

A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z

A

Attrition, sometimes called customer churn, refers to the loss of clients or customers. A churn study, then, finds these people and determines why they have ceased to patronize a business.

An awareness and usage study finds users and non-users of and measures sentiment about a product, brand or service. It enables analysts to understand who uses an offering, how and why. This study type may incorporate unaided (unprompted) as well as aided awareness elements. Some of the key metrics that emerge include market penetration, consumer ratings, typical user characteristics, and brand awareness.

B

Brand awareness describes the extent to which potential customers recognize and associate your brand with a particular product or category. Analysts express brand awareness as a percentage of your target market. It is imperative that you track and leverage your brand awareness in the period just after launching a new offering.

Brand recall refers to the ability of study respondents to remember a name (i.e. of a brand, company or product) as part of a particular class.

Brand recognition refers to the ability of study respondents to recognize a brand as associated with a particular set of attributes, such as a tagline, logo, or unique characteristics.

Brand loyalty refers to the faithfulness that a consumer feels to a particular brand and is expressed in terms of that consumer’s intent to buy, irrespective of their susceptibility to advertising by competing offerings.

Brand equity refers to the added value that your brand gives to your product. Its value is equal to the difference between sales that your branded product achieves and sales that it would achieve were it “generic,” or unbranded.

Brand equity analysis can help you enumerate a set of attributes by which to measure your brand equity and then evaluate your performance and compare your brand equity to your competitors’. A brand equity assessment can help you leverage and expand the value that your brand adds to your offering.

C

Competitive benchmarking is a process whereby businesses use primary data about their competitive set and industry best practices to pin down several core dimensions of success. Having developed a list of such “benchmarks,” a business may incorporate them into an analysis of its performance. This type of analysis could stand alone or form part of a tracking study to help the business chart growth over time.

Concept testing assesses consumer response to an idea before a product hits the shelf. It can take the form of a focus group, in which consumers react to and comment on specific features of a product idea, or a survey. Concept testing is indispensable to introducing a successful product and maintaining its success in the long term.

A confidence interval is a description of the uncertainty surrounding a given statistic. A confidence interval of +5% hence suggests that the actual value of a figure is within five percent of either side of the depicted value.

A nationally-representative consumer panel is a pre-screened pool of consumers who agree to participate in market research. This research can take the form of surveys, focus groups, or community-based studies. The panel is a stable source of consumer sample, provided that powerful techniques are employed to guarantee balance and representativeness. The FGI SmartPanel™ is an example of an online consumer panel.

Conjoint analysis, like MaxDiff, is a discrete choice model. However, rather than rating isolated attributes in relation to one another, conjoint analysis evaluates which combination of attributes influences consumer preferences. Data is typically collected through a questionnaire.

Copy testing assesses a messaging vehicle (online, on TV, in print, or in some other format) for attributes like message strength, recall, and impact on intent to buy, among others. Copy testing may include quantitative or qualitative data.

Correlation describes the extent to which changes in a dependent variable correspond to changes in an independent variable.

Correspondence analysis is one way of describing the relationship between rows and columns of categories of variables. A table of these data points could be two- or, for more complex analyses, multi-dimensional. An example is a study that differentiates between how consumers of different genders rate an array of brands. Analysts use correspondence analysis in comparing and predicting consumer behavior and preferences.

A custom panel is a pre-screened selection of a company’s customers, recruited using a database of contact information, who sign up for accounts and agree to participate in market research. In short, the panel is a stable pool of customers, available on a long term basis as the foundation for study sample. Study types that a custom panel supports include online focus groups and surveys.

A custom panel-community blends a custom panel with an online community such that the combined platforms facilitate fully-integrated qualitative and quantitative research.

D

The dependent variable is the variable in a predictive mathematical model that fluctuates in response to changes in an independent variable.

Discrete choice modeling invites respondents to choose between two or more discrete alternatives, such as products or attributes, or sets of attributes, rating these options in relation to one another rather than in isolation. MaxDiff and conjoint analysis are examples of discrete choice models.

Discriminant analysis is a statistical function that groups subjects based on a collection of predictor variables. Should a business want to introduce a product or service to a new market, an analyst could use demographic information about existing users to establish a set of predictors for who among the new group is likely to adopt. Discriminant analysis is useful in orchestrating successful product launches or the introduction of new features to an existing offering.

F

Factor analysis is a statistical method used to unearth the underlying variables, or factors, that drive change in a group of observable attributes. Analysts might use factor analysis to reduce a complex set of data to a few key attributes; for instance, they might attempt to explain consumer behavior in terms of three or four main motivations –hence why factor analysis is known as key drivers analysis in some cases.

A focus group is a research method in which a sample of individuals gives subjective feedback about a new product, service, advertisement or concept. It can be used to assess packaging, ad effectiveness, evaluate concept ideas for products or services, and any other traditional application of qualitative research.

G

Gap analysis is a statistical method that estimates the space between a company’s actual performance and its desired performance and may apply to metrics like sales, market penetration, brand awareness, loyalty, customer satisfaction, and others. The objective of gap analysis in marketing science is to quantify the gap between where a company is and where it wants to be and in so doing create the potential for a step-by-step process to achieve desired results.

I

iGAGE is the patented sample-balancing method that FGI Research uses to make sure that each sample pulled from our consumer panel, SmartPanel, is nationally-representative. It weights responses according to consumer income, gender, age, geographic location and ethnicity so that outcomes reflect demographics as documented by the latest U.S. Census.

Incidence refers to the frequency with which a particular event occurs. Often used in medical analysis, the term can also describe market trends; for instance, how often people within a certain population bought or used a particular product or service.

An independent variable is the variable in a predictive mathematical model which influences the dependent variable.

L

Loyalty, in the context of marketing, refers to a consumer’s commitment to continue using a service, product or brand. Another indicator of loyalty is word-of-mouth advertising, or the frequency and level of enthusiasm with which a user would recommend a product, service or brand to acquaintances in daily conversation and on social media.

M

Market research is an organized effort to gather intelligence about markets, customers, companies and industries through the use of qualitative and quantitative methodologies. Businesses may use the metrics studied in market research to inform brand management, positioning, customer service, and marketing strategies.

A market research online community is a static platform to which a company invites its customers to engage in online discussion. The platform supports a range of emerging qualitative research methods by generating topic-driven feedback and original insight from a company’s customers. The FGI Research online community enables members to contribute to any number and variety of one-to-two-week-long forum discussions, each paired with a quantitative poll.

MaxDiff, also known as maximum difference scaling, is a discrete choice model used to evaluate products or services in relation to one another. Compared to standard rating scales, which assign a numeric rating to a product in isolation, MaxDiff offers more actionable results. It enables a business to directly assess how its offering compares to competing offerings, rather than acting on a vague sense of how each rates individually. The method is especially useful in brand and product positioning.

Mixed-Mode Optimization is a process whereby FGI Research generates an optimal mixture of modalities (for instance, phone and web, phone and mobile, or community and survey) to support your research needs.

Modality refers to the platform used to collect the data in market research. Examples include phone, web, mobile devices and mail. Read more about FGI’s data-collection capabilities here.

Multivariate regression expands the regression function to describe not just one dependent and independent variable but multiple variables. It is especially useful in developing predictive mathematical models, which can be used to anticipate actual events.

O

An online focus group, whether in bulletin-board or chatroom format, may incorporate interactive media to facilitate targeted feedback on specific elements of a concept, messaging vehicle or other object of inquiry. This platform adapts the original concept of a focus group to an online format, reducing costs and logistics and expanding opportunities for insight. Learn more about online focus groups.

P

Packaging market research uses quantitative and, sometimes, qualitative data to gauge how consumers react to product packaging or packaging concepts. An analyst may assess consumer packaging preferences with regards to attributes like weight, color, size, and ease of use, and from there make suggestions for a successful product launch or relaunch.

Positioning is the process whereby marketers develop an identity for a service, product or brand, and leverage that identity to create points of comparison with competitors and increase the service’s, product’s, or brand’s standing in the market.

Projectability describes the extent to which results from a sample population are true of the general population. In other words, can the statistics from a given study be “projected” on all individuals within the category on which that study focuses? Projectability depends on the reliability of the sampling and weighting methods used in the survey process.

Q

Qualitative research generates verbatim responses to questions about a particular topic or set of topics using methods like focus groups and online communities. These subjective responses can be used to create a picture of consumer sentiment, place quantitative data in context, refine impending quantitative studies and unearth new ideas.

Quantitative research, using surveys, generates numeric data reflective of respondents’ opinions on a particular topic or set of topics. The data can be used in analysis and reporting to help inform business decisions. Survey fieldwork may be completed online or through the use of traditional telephone interviewing.

R

Random digit-dialing is the process whereby researchers guarantee sample representativeness in a phone study. Numbers are dialled randomly by an automatic dialer to guard against bias and generate a sample representative of a given population.

A regression function describes the relationship between a dependent variable, or outcome; and independent variables, which may be manipulated in order to affect that outcome.

S

A study sample is a group of consumers or customers, whose size varies, selected to participate in a given market research study. A sample can be customized to reflect age, gender, geographic, income, preferential or ethnic requirements, among others, using segmentation.

Satisfaction testing enables businesses to measure their customers’ needs from, feelings about, and level of satisfaction with a product or service –at a single moment, or across a period of time. Keeping tabs on satisfaction helps companies identify and satisfy unmet needs, as well as find points for growth in their current approach.

Segmentation divides your market into customer bases according to attributes like sex, race, age, income and geography. Purchase and behavior data can be added to segmentation scoring to enable you to further understand your audience in terms of values and preferences.

SmartPanel™ is FGI’s general-population consumer panel of U.S. consumers. Using surveys, topic voting, quick polls and an integrated online community platform, SmartPanel assesses consumer opinion on a variety of topics. It is the source for FGI Research TrendTruth™ studies as well as an optimal source of study sample for client-owned consumer research studies.

A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z