Market Research and Consumer Insights

01Purpose and scope

• This manual supports the course Market Research and Consumer Insights. It describes how the work is actually done in marketing teams, agencies and insight groups, from the first request for information through to the decisions that follow. It is written as a reference for practitioners who commission studies, run them, or act on what they find.
• Scope covers primary research such as surveys, depth interviews, focus groups, usability sessions and concept tests, along with secondary sources such as sales records, panel data, syndicated reports and web analytics. It does not teach statistics in depth and it does not replace formal methodology training. Where a decision carries legal, financial or health consequences, this guidance does not replace qualified professional judgment, approved procedures or required authorization.
• The same practices apply to consumer products, services, retail, media and nonprofit campaigns, but the emphasis shifts with the setting. A subscription software company may need churn reasons and feature usage patterns, while a packaged food brand may care more about shelf choice, trial and purchase frequency.

02Core principles

• Research exists to inform a decision. Before anything is fielded, the team should be able to name the decision, the options on the table, and the kind of evidence that would change the choice. A study that cannot change anything is a cost without a return, even when the findings are interesting.
• Consumer insight is different from consumer opinion. A stated preference in a survey is a signal, and observed behavior such as purchase records, repeat visits or abandoned carts is often a stronger one. Experienced practitioners compare the two and say plainly where they disagree.
• Respect for participants and for the data they provide is not optional. Consent, a clear statement of purpose, collecting only what is needed and storing material securely come before convenience. Participants who feel misled tend to give worse answers and damage the reputation of the research function.

03Planning and preparation

• Start with a written research brief of one or two pages. It should state the business question, the decision owner, the deadline, the budget range, the target audience, what the team already knows, and what is explicitly out of scope. Briefs that skip the existing knowledge section often commission studies that repeat a finding from last quarter.
• Choose the method by what the decision needs. Quantitative surveys give sizes, segments and trends but cannot explain motives in depth. Qualitative interviews explain reasons and language but cannot project results to a market. Behavioral data is large and current but silent on why people act. Many projects combine two of these, for example interviews to shape survey wording and then a survey to size what the interviews uncovered.
• Define the population before recruiting. For a survey of households that buy a product monthly, specify the purchase frequency, age range, region and whether lapsed buyers count. Decide where the sample will come from: an opt-in panel, intercepts at a store, a customer list, or a recruitment agency. Each source carries its own bias, and the write-up should name it. Quotas on age, gender, region or usage keep the sample close to the population, but they do not correct for people who choose to respond.
• Identify stakeholders early. Brand managers and category leads usually own the decision. Sales and customer service often hold context that changes how findings are read. Finance may need to approve spend, and privacy or legal staff need to review consent language and data handling before fieldwork begins. Agencies and vendors should receive the brief, the timeline and the data rules in writing.

04Recommended workflow

1. 1
   : Write the research question as a decision statement. Replace a vague goal such as understand millennials with a specific choice such as whether to launch a smaller pack size in two regions. Confirm the decision owner has signed off on the question and the evidence threshold.
2. 2
   : Review what already exists. Check internal sales data, past studies, customer service logs, app store reviews and available syndicated reports. Record what was found, the date it was collected, and any reason it may not apply to the current question. This step often shortens the new work.
3. 3
   : Select the method and design the instrument. For surveys, draft questions that are neutral, one idea each, and ordered from general to specific. For interviews, prepare a discussion guide with open prompts and a few planned probes. Keep the instrument as short as the decision allows, because length increases dropout and reduces answer quality.
4. 4
   : Pilot the instrument. Run five to ten test participants or a small soft launch. Time the survey, check that skip logic works on phones and desktops, and look for questions that people read differently than intended. Revise before the main fieldwork starts, not after the first wave of responses arrives.
5. 5
   : Field the study and monitor it. Track quota progress daily, watch completion times, review open-ended answers for patterns of straight-lining or copy-paste responses, and keep a log of any changes made mid-field. If a recruitment source underperforms, decide openly whether to supplement it and document the change.
6. 6
   : Clean, analyze and test the findings. Remove duplicates, speeders and failed attention checks according to rules written before fieldwork. Run the core tables or coded themes, then check whether results hold when segments are split or when behavioral data is compared with what respondents reported. Note any weighting and explain it in plain terms.
7. 7
   : Write the readout around the decision. Lead with the answer to the question, state the confidence and limits, show the evidence in a few charts or quotes, and end with the implications for the choice on the table. Avoid producing a deck of every table. Schedule the readout with the decision owner so the findings are discussed while the choice is still open.

05Quality checks, risk controls and compliance

• Quality starts before fieldwork. Review the questionnaire against the brief for leading wording, double-barreled items and answer lists that do not cover the likely responses. Ask someone outside the project to read it cold. For qualitative work, check that the moderator guide does not presuppose the conclusion.
• During fieldwork, use attention checks, consistency items and time thresholds, but set them thoughtfully. A single failed attention check can be a careless respondent or a confusing question. Document the rule, apply it consistently, and report how many responses were excluded and why. For probability samples, a margin of error can be stated; for opt-in panels and intercept samples it is not a valid measure of precision, and reporting it as one misleads readers.
• Privacy and consent rules depend on jurisdiction and change over time. In the United States, state consumer privacy laws vary in their scope and requirements. The European Union and the United Kingdom apply their own data protection regimes, which govern personal data collection, lawful basis, retention and transfers. Check the current text from the relevant regulator or official legal source, and involve your privacy or legal team before recording interviews, collecting contact details or linking survey answers to customer records.
• Several specific controls are common. Obtain explicit consent before audio or video recording and say who will see the material. Remove direct identifiers from transcripts and datasets before sharing them outside the core team. Set a retention period and delete raw data when it expires. Confirm rules on research with children, which often require parental consent and vary by location. Review incentive practices against platform terms and local rules on prize draws and payments.
• Professional codes are useful reference points. The ESOMAR International Code of Conduct for market, opinion and social research is widely used by practitioners and vendors, and it addresses transparency, participant protection and client confidentiality. Adopt its principles into your own process and check the current version rather than relying on memory. Where a vendor or agency handles fieldwork, require written confirmation of their compliance and data handling practices.

06Collaboration and documentation

• Research succeeds when the people who will use it are involved at the right moments: the brief, the instrument review, a mid-field check on sample and early themes, and the readout. Keep a shared project folder with the brief, approved instrument, sampling plan, field notes, data dictionary, analysis plan and final report. A reader six months later should be able to see what was asked, of whom, and how the results were produced.
• Write a short methodology note for every study. It should cover the method, the sample source and size, the fieldwork dates in the project file rather than in the report body if your organization prefers, the exclusion rules, any weighting, and known limitations. Keep it to a page. Stakeholders read the summary, but auditors and future researchers read the note.
• Define roles explicitly. A typical project has a decision owner, a project lead who manages scope and timeline, a research specialist who designs the method and analysis, a fieldwork partner, a data or privacy reviewer, and a presenter who carries the findings to the business. In small teams one person may hold several roles, but the responsibilities should still be named in writing.
• Manage disagreement openly. When a finding conflicts with a strong internal belief, present the evidence, the limits and a proposed way to test it further. Record the discussion and the decision that followed. This protects the research function from being used selectively to support a conclusion already reached.

07Measures and continuous improvement

• Measure the research function as well as individual studies. Useful indicators include the share of studies that name a decision owner and a decision in the brief, the percentage of findings that led to a documented action within a set period, the time from brief to readout, field quality metrics such as exclusion rates, and the proportion of projects that reused existing data before commissioning new work. Track these over several projects, because a single study tells you little about process health.
• Ask decision owners for a short review after each major study. Useful questions include whether the findings changed the choice, whether anything was missing from the readout, and whether the timeline matched the decision window. Keep the answers in a simple log so patterns are visible. Repeated complaints about late readouts, for example, usually point to brief approval delays rather than analysis speed.
• Compare predictions with outcomes where you can. If a concept test predicted strong trial for a product and the launch achieved little trial, review the sample, the stimulus and the market conditions. Not every gap is a research failure, but recording these comparisons builds a more honest picture of what the methods can and cannot forecast.
• Limits should be stated plainly. Consumer research has trouble with rare purchases, with categories where people cannot articulate their choices, and with fast changing trends where the sample is out of date before the report is finished. In those cases, a mix of small qualitative work, monitoring of behavioral signals and a willingness to run small in-market tests may serve better than a large survey.

08Common mistakes and how to prevent them

• Mistake one is starting with a method rather than a question. Teams often ask for a survey because the last one worked, then discover it cannot answer the actual decision. Prevent this by requiring the decision statement in the brief before any vendor is contacted.
• Mistake two is using a convenience sample and describing the results as representative of all consumers. The fix is to name the sample source and its limits in the headline and the methodology note, and to avoid population claims unless the design supports them.
• Mistake three is adding questions to an instrument after it has been approved. Each addition lengthens the survey, raises dropout and often introduces a question that was not piloted. Keep a parking lot for new ideas and move them into the next study.
• Mistake four is reporting significance or differences that are too small to matter. A statistically detectable gap of one point may have no business consequence. Agree in advance on the size of difference that would change a decision, and report results relative to that threshold.
• Mistake five is treating a readout as a deliverable rather than a conversation. Slides that are never discussed with the decision owner rarely change anything. Book the discussion before the report is final and capture the decisions that come out of it.

09Practical checklist

• Write the research question as a decision statement and confirm the decision owner has approved it.
• Review existing internal data, past studies and syndicated sources before commissioning new work.
• Choose the method by what the decision needs, and name its strengths and limits in the brief.
• Define the target population, the sample source, quotas and exclusion rules before fieldwork starts.
• Pilot the instrument with a small group and revise wording, length and logic before launch.
• Confirm consent, recording permissions, data retention and privacy review with the relevant team.
• Monitor quotas, completion times and response quality during fieldwork and log every mid-field change.
• Apply exclusion and weighting rules as written in advance and report how many responses were removed.
• Write a one page methodology note covering method, sample, rules and limitations.
• Lead the readout with the answer, the confidence level and the implication for the decision.
• Schedule a post study review with the decision owner and record what changed as a result.

★Learning sources

• Micro Degree: Market Research and Consumer Insights
• Micro Degree: Marketing Strategy Development
• Micro Degree: Programmatic Advertising Management
• Micro Degree: Lead Generation Operations