Marketing research helps organizations understand markets, audiences, customers, competitors, products, communication, and business opportunities through systematic collection and analysis of information. The AAMA Marketing Research Methods Guide provides a practical overview of commonly used research approaches and explains when different methods are most appropriate.
This guide is designed for marketing and advertising professionals, agencies, brands, students, academics, researchers, entrepreneurs, and organizations making decisions that require evidence rather than assumption. Research does not need to be unnecessarily complicated, but the method should match the question being asked, the type of evidence required, and the limitations of the available data.
Start With the Research Question
Every research project should begin with a clearly defined question. Selecting a method before defining the question can lead to collecting large amounts of information that never actually address the decision the organization needs to make.
Useful research questions might include:
- Why are customers choosing one product over another?
- Which audience segments are most likely to respond to a new offering?
- How well is a brand recognized within its target market?
- What prevents prospects from completing a purchase?
- Which message is easiest to understand?
- How do customers describe the problem a product is intended to solve?
- What features matter most during consideration?
- How is customer satisfaction changing over time?
- What market opportunities are emerging?
- How does advertising affect awareness or purchase intent?
A strong research question gives the project direction and makes it easier to determine what information should be collected.
Define the Research Objective
The research objective explains what the organization needs to learn and how that knowledge will support a decision. It should translate a broad business concern into a specific information need.
For example, a business problem might be:
“Sales of the premium product have declined.”
The research objective might be:
“Determine whether the decline is associated with price perceptions, competitive alternatives, product satisfaction, or changes in customer needs.”
That objective provides clearer direction for choosing research methods.
Distinguish Exploratory, Descriptive & Causal Research
Marketing research can be broadly organized according to the type of question being investigated. Understanding these differences helps prevent organizations from expecting one type of research to answer questions it was not designed to address.
Exploratory Research
Exploratory research is used when the organization needs to better understand a problem, audience, behavior, or opportunity before developing more specific hypotheses.
It is useful for questions such as:
- What might be happening?
- Why might customers behave this way?
- What language do customers use?
- Which issues should we investigate further?
- What possibilities have we overlooked?
Interviews, focus groups, observational research, secondary research, and open-ended analysis are commonly used for exploratory work.
Descriptive Research
Descriptive research measures or describes characteristics, behaviors, attitudes, preferences, or conditions within a population or market.
It is useful for questions such as:
- What percentage of customers recognize the brand?
- How often do customers purchase the product?
- Which channels do customers use?
- What features do people prefer?
- How satisfied are customers?
Surveys, panels, market data, customer databases, and observational measurement are commonly used for descriptive research.
Causal Research
Causal research attempts to determine whether changing one factor causes a change in another. Experiments and controlled tests are commonly used when causal inference is the objective.
It is useful for questions such as:
- Did the new advertisement increase conversions?
- Does a lower price increase purchase rate?
- Did the redesigned landing page cause more registrations?
- Does message A perform better than message B?
Causal conclusions generally require stronger research design than simple observation or correlation.
Understand Qualitative Research
Qualitative research focuses on understanding experiences, motivations, language, perceptions, attitudes, and reasoning in depth. It usually works with smaller samples and produces detailed information rather than population estimates.
Qualitative methods are particularly useful when researchers need to understand why something is happening or discover issues they did not know to ask about in advance.
Common qualitative methods include:
- Individual interviews
- Focus groups
- Observation
- Ethnographic research
- Diary studies
- Open-ended responses
- Online communities
- Qualitative usability testing
Qualitative findings can generate hypotheses, identify language, reveal motivations, and improve the design of later quantitative research.
Understand Quantitative Research
Quantitative research uses numerical data to measure characteristics, behaviors, attitudes, relationships, or outcomes. It is commonly used when organizations need estimates, comparisons, trends, statistical analysis, or evidence from larger samples.
Common quantitative methods include:
- Surveys
- Experiments
- Customer-data analysis
- Sales analysis
- Web analytics
- Advertising measurement
- Market datasets
- Panels
- Structured observation
Quantitative research can help answer questions about how much, how often, how many, or how strongly variables are related.
Primary Research
Primary research involves collecting new information specifically for the research question being investigated. The organization controls the research design, sample, questions, and collection process.
Examples include:
- Customer surveys
- Interviews
- Focus groups
- Experiments
- Observational studies
- Product tests
- Concept tests
- Usability research
Primary research can provide highly relevant information, but it usually requires more time, planning, and resources than using existing data.
Secondary Research
Secondary research uses information that has already been collected by another organization or for another purpose. It can provide valuable market context before new primary research is commissioned.
Sources may include:
- Government statistics
- Academic research
- Industry reports
- Trade publications
- Company reports
- Market databases
- Existing customer research
- Historical campaign data
- Public financial information
- Advertising archives
Secondary research is often the best first step because existing information may answer part of the question or reveal what additional research is actually necessary.
Surveys
Surveys collect structured responses from participants using a standardized set of questions. They are one of the most commonly used marketing research methods because they can measure attitudes, behaviors, awareness, satisfaction, preferences, and other characteristics across relatively large samples.
Surveys are useful when researchers need to estimate:
- Brand awareness
- Customer satisfaction
- Purchase intent
- Product usage
- Media behavior
- Preferences
- Demographic characteristics
- Attitudes
- Consideration
- Reported behavior
Survey results depend heavily on question wording, sample quality, response rates, and research design.
Related AAMA Resource: How to Design a Marketing Survey
Individual Interviews
Individual interviews are one-on-one conversations between a researcher and participant. They allow the researcher to explore experiences, motivations, decision processes, attitudes, and language in greater depth than most structured surveys.
Interviews are especially useful when:
- The subject is complex
- Participants have specialized knowledge
- Individual experiences matter
- Sensitive topics may be difficult to discuss in a group
- Researchers need to ask follow-up questions
- The decision process needs to be understood in detail
A skilled interviewer follows the research objectives while allowing unexpected but relevant insights to emerge.
Focus Groups
Focus groups are moderated discussions involving a small group of participants selected according to defined research criteria. They allow researchers to observe how people discuss ideas, react to concepts, and respond to the perspectives of others.
Focus groups can be useful for:
- Exploring attitudes
- Reviewing early concepts
- Understanding language
- Identifying concerns
- Generating hypotheses
- Exploring category perceptions
Focus groups should not be used to estimate how common an opinion is across an entire market. The small and non-random sample generally makes population-level conclusions inappropriate.
Observational Research
Observational research studies what people actually do rather than relying only on what they report doing. Researchers may observe behavior in stores, workplaces, homes, websites, applications, events, or other environments.
Observation can reveal:
- Shopping behavior
- Product usage
- Navigation patterns
- Physical interactions
- Workarounds
- Points of confusion
- Environmental influences
People are not always able to accurately explain their own behavior, so observation can provide insights that interviews and surveys may miss.
Ethnographic Research
Ethnographic research examines behavior, routines, environments, and cultural context in greater depth. It often involves observing participants in real-world settings and may include interviews, artifacts, photographs, diaries, or other contextual information.
Ethnographic approaches can be particularly valuable when products or services are closely connected to daily routines, social relationships, work practices, or cultural behavior.
The goal is to understand behavior within its actual context rather than isolating it in an artificial research setting.
Diary Studies
Diary studies ask participants to record experiences, behaviors, thoughts, or activities over a period of time. Entries may be collected through written diaries, mobile applications, photographs, videos, or structured prompts.
Diary research can be useful for understanding:
- Recurring behaviors
- Long purchase journeys
- Product usage over time
- Changing attitudes
- Daily routines
- Customer experiences
- Media consumption
The method allows researchers to capture experiences closer to when they occur rather than asking participants to reconstruct them much later.
Customer Feedback Research
Customer feedback research uses information collected from existing customers about their experiences with products, services, support, or the organization.
Sources may include:
- Satisfaction surveys
- Reviews
- Support interactions
- Interviews
- Complaints
- Product feedback
- Customer-service records
- Online communities
- Cancellation reasons
Customer feedback can identify recurring problems and opportunities, but researchers should remember that vocal customers may not represent the entire customer population.
Social Listening
Social listening examines publicly available online conversations, mentions, discussions, and reactions related to brands, products, competitors, industries, or topics.
It can help identify:
- Emerging concerns
- Audience language
- Brand perceptions
- Competitive discussion
- Cultural trends
- Customer-service issues
- Campaign reactions
Social data should be interpreted carefully because people posting publicly are not necessarily representative of the broader target audience.
Search Behavior Research
Search behavior can provide information about what people are actively trying to learn, compare, find, or purchase. Search queries can reveal language, questions, needs, category interest, and changing demand.
Researchers may examine:
- Search-query data
- Search trends
- Website search
- Paid-search data
- Search-engine results
- Keyword research
Search behavior can provide valuable evidence of expressed interest, but search volume alone does not explain the motivations behind that behavior.
Website & Digital Analytics
Digital analytics measure user behavior across websites, applications, campaigns, and other digital environments. These data can provide direct evidence of what users do after arriving rather than relying entirely on self-reported behavior.
Common measures include:
- Users
- Sessions
- Traffic sources
- Pageviews
- Engagement
- Conversion rate
- Path behavior
- Ecommerce activity
- Campaign performance
Analytics can show what happened but may not explain why it happened. Combining behavioral data with qualitative research can provide a more complete understanding.
Customer & CRM Data
Customer databases and CRM systems can provide information about purchasing, retention, lead progression, customer value, sales activity, and relationship history.
Research using customer data can help identify:
- High-value segments
- Purchase frequency
- Retention patterns
- Churn
- Cross-selling opportunities
- Geographic patterns
- Lead quality
- Customer lifecycle behavior
Researchers should consider data quality, missing information, privacy requirements, and whether the available data accurately represents the question being investigated.
Sales Data
Sales data can reveal changes in demand, product performance, geographic variation, seasonality, customer behavior, and campaign outcomes.
Researchers may analyze:
- Revenue
- Units sold
- Average order value
- Product mix
- Purchase frequency
- Geographic performance
- Channel performance
- Promotional response
Sales data provides important evidence of actual behavior, but changes in sales can have multiple causes. Researchers should avoid assuming that marketing activity caused a change without additional evidence.
Competitive Research
Competitive research examines the organizations, products, services, pricing, positioning, communication, customer experiences, and activities within a competitive environment.
Potential sources include:
- Competitor websites
- Advertising
- Public pricing
- Product reviews
- Annual reports
- Retail environments
- Search visibility
- Social media
- Public interviews
- Industry publications
Competitive research should focus on information that supports strategic decisions rather than simply documenting everything competitors do.
Advertising Research
Advertising research evaluates creative concepts, messages, media, awareness, response, and campaign effectiveness.
Methods can include:
- Concept testing
- Copy testing
- Brand-lift studies
- Recall studies
- Recognition studies
- Experiments
- A/B testing
- Surveys
- Eye-tracking
- Qualitative interviews
Advertising research can occur before, during, or after campaign execution depending on what the organization needs to learn.
Concept Testing
Concept testing evaluates audience reactions to early product, service, campaign, positioning, or creative ideas before full development.
Researchers may evaluate:
- Relevance
- Clarity
- Appeal
- Differentiation
- Credibility
- Purchase interest
- Confusion
- Objections
Concept testing can help identify problems early, but it should not turn creative development into a popularity contest. Research should evaluate whether a concept accomplishes its strategic purpose rather than simply asking participants which idea they like best.
Copy Testing
Copy testing evaluates advertising messages or creative executions before or during a campaign. Testing can assess comprehension, recall, persuasion, relevance, credibility, or behavioral response.
Methods may include:
- Surveys
- Interviews
- Experiments
- A/B testing
- Recall testing
- Recognition testing
- Digital performance testing
The appropriate method depends on what aspect of the communication needs to be evaluated.
A/B Testing
A/B testing compares two versions of a marketing element to determine whether one produces a stronger result against a defined measure.
Possible test elements include:
- Headlines
- Calls to action
- Images
- Advertisements
- Landing pages
- Email subject lines
- Offers
- Page layouts
The strongest tests isolate the variable being evaluated so differences can be interpreted more confidently.
Related AAMA Resource: A/B Testing Guide for Marketers
Experimental Research
Experiments manipulate one or more variables while attempting to control other factors that might influence the outcome. They are particularly useful when researchers want to make causal conclusions.
Marketing experiments may involve:
- Pricing
- Creative
- Promotions
- Landing pages
- Media exposure
- Offers
- Product features
- Messaging
Random assignment, control groups, and careful measurement can strengthen causal interpretation.
Test Markets
Test markets introduce a product, campaign, price, promotion, or other marketing activity within a limited market before wider rollout. The test can provide evidence about customer response and operational requirements under more realistic conditions.
Test markets may be defined by:
- Geography
- Stores
- Customer groups
- Digital audiences
- Channels
Researchers should consider whether the test market accurately represents the conditions expected during broader deployment.
Usability Research
Usability research examines how effectively people can use websites, applications, products, interfaces, or other systems. Participants may be asked to complete tasks while researchers observe where they encounter confusion or difficulty.
Usability research can identify:
- Navigation problems
- Unclear instructions
- Form friction
- Interface confusion
- Information gaps
- Accessibility issues
- Conversion barriers
The purpose is usually to improve the experience rather than measure general market attitudes.
Brand Tracking Research
Brand tracking measures brand-related perceptions and attitudes repeatedly over time. It can help organizations understand whether awareness, consideration, preference, perceptions, or other brand measures are changing.
Common tracking measures include:
- Unaided awareness
- Aided awareness
- Consideration
- Preference
- Purchase intent
- Brand attributes
- Familiarity
- Usage
Consistency in methodology is especially important because the value of tracking comes from comparing results over time.
Customer Satisfaction Research
Customer satisfaction research measures how customers evaluate their experiences with a product, service, interaction, or organization.
Common measures can include:
- Customer satisfaction
- Net Promoter Score
- Customer effort
- Product satisfaction
- Service satisfaction
- Likelihood to repurchase
Satisfaction scores become more useful when organizations examine what drives those scores and how they relate to actual customer behavior.
Segmentation Research
Segmentation research identifies meaningful groups within a broader market or customer base. Segments may be based on behavior, needs, attitudes, value, demographics, geography, purchase patterns, or combinations of variables.
Effective segments should be:
- Meaningfully different
- Relevant to marketing decisions
- Identifiable
- Reachable
- Large or valuable enough to matter
- Actionable
Creating elaborate segments that cannot influence products, messaging, media, or service decisions provides little practical value.
Pricing Research
Pricing research examines how customers perceive prices, tradeoffs, value, and willingness to pay. It can help organizations understand how different pricing structures might influence demand or positioning.
Methods may include:
- Surveys
- Experiments
- Conjoint analysis
- Choice modeling
- Sales analysis
- Market testing
Stated willingness to pay may differ from actual purchasing behavior, so pricing research is often stronger when several evidence sources are combined.
Conjoint Analysis
Conjoint analysis is a quantitative research technique used to estimate how people value different combinations of product or service attributes. Participants make choices among alternatives, allowing researchers to infer the relative importance of characteristics such as price, features, brand, service, or configuration.
Conjoint methods can be useful for product development, pricing, packaging, and feature prioritization, but they require careful study design and analysis.
Panel Research
A research panel is a group of participants who have agreed to participate in research over time or who can be sampled for specific studies. Panels can provide efficient access to defined populations and allow repeated measurement.
Researchers should understand how panel participants were recruited and whether repeated participation may influence behavior or responses.
Longitudinal Research
Longitudinal research studies the same people, customers, groups, or measures across multiple points in time. It can help researchers understand change, retention, development, and longer-term relationships.
Longitudinal studies may be useful for:
- Brand tracking
- Customer behavior
- Product adoption
- Satisfaction
- Retention
- Attitude change
Participant attrition can create challenges when studies continue over long periods.
Cross-Sectional Research
Cross-sectional research collects information from a population at a particular point in time. Many marketing surveys and market studies use this approach.
Cross-sectional research is useful for describing current conditions, attitudes, or behaviors, but it is less effective for determining how individual participants change over time.
Sample Selection
The sample is the group of people, organizations, transactions, or observations included in a study. A good sample should reflect the population relevant to the research question.
Researchers should consider:
- Who qualifies to participate?
- Who is excluded?
- How are participants recruited?
- Is the sample representative?
- Are important groups underrepresented?
- Is the sample large enough for the intended analysis?
A large sample does not automatically correct a biased sampling process.
Probability Sampling
Probability sampling gives members of the target population a known chance of selection. Methods can include simple random sampling, stratified sampling, systematic sampling, and cluster sampling.
Probability methods can support stronger population estimates when a suitable sampling frame exists.
They are not always practical in commercial marketing research, but researchers should understand the tradeoffs when non-probability methods are used instead.
Non-Probability Sampling
Non-probability sampling does not give every member of the population a known chance of selection. Common examples include convenience samples, volunteer samples, quota samples, customer lists, and online panels.
These methods are widely used because they can be practical and cost-effective, but researchers should be cautious when generalizing results to a broader population.
Sample Size
Sample size affects the precision and analytical possibilities of quantitative research. The appropriate size depends on the population, sampling method, desired precision, expected variation, subgroup analysis, available resources, and study objectives.
More responses are not automatically better if the sample is poorly selected. Sample quality and research design matter alongside sample size.
Research Bias
Bias is a systematic influence that can distort research results. Bias can enter through sampling, question wording, interviewer behavior, participant selection, measurement, analysis, or interpretation.
Common forms include:
- Selection bias
- Response bias
- Nonresponse bias
- Social-desirability bias
- Confirmation bias
- Recall bias
- Interviewer bias
- Measurement bias
Researchers should attempt to identify likely sources of bias before collecting data and acknowledge meaningful limitations when reporting results.
Correlation Is Not Causation
Two variables can move together without one causing the other. A correlation may result from coincidence, a third factor, reverse causation, or another underlying relationship.
For example, a campaign may receive more advertising impressions during the same period that sales increase, but the relationship alone does not prove that the advertising caused the sales increase.
Causal claims generally require stronger evidence, such as controlled experiments or carefully designed quasi-experimental methods.
Combine Research Methods
Many marketing questions benefit from using several methods rather than relying on a single source of evidence.
For example, an organization investigating customer churn might combine:
- CRM data to identify when customers leave.
- Surveys to measure reported reasons.
- Interviews to explore those reasons in depth.
- Product analytics to identify behavioral patterns.
- Experiments to test proposed improvements.
Different methods can answer different parts of the same question.
Use Qualitative Research Before Quantitative Research
When an organization does not understand the problem well, beginning with qualitative research can improve later quantitative studies. Interviews or focus groups may reveal terminology, motivations, barriers, and unexpected issues that researchers can then measure more systematically.
This sequence can prevent survey designers from forcing participants to choose among answer options that do not reflect how customers actually think about the issue.
Use Quantitative Research to Test Scale
Qualitative research can reveal an important idea without showing how common that idea is. Quantitative research can then help determine whether the observation is widespread, limited to a particular segment, or relatively uncommon.
The two approaches are complementary rather than competing.
Separate What People Say From What They Do
Self-reported research and behavioral data measure different things. A customer may report that price is the most important factor while actual purchasing behavior shows a stronger response to convenience, availability, or brand familiarity.
Neither form of evidence should automatically be treated as superior. Differences between stated attitudes and observed behavior can themselves provide useful insight.
Protect Research Participants
Research involving people should be conducted responsibly. Participants should understand what is being asked of them, and organizations should use appropriate practices for consent, privacy, data collection, storage, and disclosure.
Research involving sensitive information, vulnerable populations, regulated subjects, academic studies, or other higher-risk situations may require additional ethical, legal, institutional, or professional review.
The need for useful marketing information does not override responsibility to participants.
Document the Methodology
Research findings are easier to evaluate when the methodology is clearly documented.
A useful methodology section can identify:
- Research objective
- Method
- Target population
- Sample
- Recruitment approach
- Field dates
- Sample size
- Questionnaire or discussion structure
- Data sources
- Analysis method
- Important limitations
Methodological transparency helps readers understand what conclusions the evidence can reasonably support.
Report Limitations
Every research project has limitations. Good research reporting identifies important constraints rather than presenting results with greater certainty than the evidence supports.
Limitations may include:
- Small sample
- Non-representative sample
- Low response rate
- Self-reported behavior
- Missing data
- Limited geography
- Short field period
- Measurement error
- Platform restrictions
- Uncontrolled variables
Acknowledging limitations strengthens the credibility of the analysis because it clarifies how the findings should be interpreted.
Separate Findings From Interpretation
Research reports should distinguish between what the data directly show and what the researcher believes those findings may mean.
For example:
Finding: 58% of respondents reported that setup appeared difficult.
Interpretation: Perceived complexity may be discouraging some prospects from progressing further in the purchase process.
The second statement is a reasonable interpretation, but it goes beyond the direct measurement.
Maintaining this distinction helps prevent research findings from becoming overstated.
Translate Research Into Decisions
Research creates value when it helps organizations make better decisions. The final stage of a research project should connect findings with the marketing, product, customer, media, or business questions that originally motivated the study.
Potential implications might include:
- Change positioning
- Refine audience strategy
- Modify product features
- Test new pricing
- Improve onboarding
- Adjust media
- Develop new messages
- Address customer-service problems
- Conduct additional research
Avoid forcing every finding into a recommendation. Some results may simply reduce uncertainty or identify questions that require further investigation.
Recommended Marketing Research Process
A practical research project can follow this sequence:
- Define the Business Problem
- Establish the Research Question
- Define the Research Objective
- Review Existing Information
- Select the Research Method
- Define the Target Population
- Develop the Sample
- Design the Research Instrument
- Test the Instrument
- Collect the Data
- Clean and Organize the Data
- Analyze the Findings
- Evaluate Limitations
- Interpret the Results
- Develop Implications
- Report the Findings
- Preserve Useful Research for Future Reference
The exact process will vary depending on the scope and method, but research should always maintain a clear connection between the original question and the evidence collected.
Choosing the Right Research Method
The best method depends on what the organization needs to know.
If the question is “Why are customers leaving?”, interviews may help uncover motivations while customer data identifies behavioral patterns.
If the question is “How many customers are satisfied?”, a well-designed survey may be appropriate.
If the question is “Which landing page produces more registrations?”, an experiment or A/B test may provide stronger evidence.
If the question is “How are customers actually using the product?”, observation or product analytics may be more useful than asking people to describe their behavior from memory.
Research design begins with matching the method to the question.
Research Should Reduce Uncertainty
Marketing research rarely eliminates uncertainty completely. Markets change, people behave unpredictably, samples are imperfect, and evidence can support several interpretations.
The purpose of research is to replace unsupported assumptions with better evidence. Good research helps organizations understand what they know, what they do not know, and which conclusions are strong enough to support a decision.
Related AAMA Resources
Continue exploring research and measurement with How to Design a Marketing Survey, A/B Testing Guide for Marketers, Campaign Measurement Framework, Marketing Metrics & KPI Reference, Marketing & Advertising Glossary, Positioning & Messaging Framework, and Common Marketing Formulas. These resources provide additional guidance for designing studies, measuring outcomes, understanding audiences, and translating evidence into marketing decisions.
The AAMA Resource Library will continue expanding with research worksheets, measurement templates, survey resources, calculators, and professional references designed to support evidence-based advertising and marketing.