Research Your Market and Customers to Help Make a Specific Business Decision

## Start by Naming the Decision the Research Must Help Make Begin with the decision the research must improve. A broad request to “understand the market” produces scattered facts, while a defined choice clarifies which people, behaviors, alternatives, and outcomes deserve attention. - Name the decision owner, deadline, available options, and cost of a wrong choice. - Convert the business concern into answerable research questions. - Record current beliefs and the evidence that would confirm, revise, or reject them. - Define what a useful answer must enable the team to decide or change. ## Set the Research Scope and Rules for Consent and Privacy Bound the study before selecting methods. Specify the population, geography, time period, product context, and exclusions, then establish quality and ethical rules for consent, privacy, access, retention, and appropriate use. - Separate facts needed for the decision from interesting background information. - Define acceptable evidence quality, uncertainty, budget, and completion criteria. - Collect the least sensitive information that can answer the question. - Document limitations that could change how findings are interpreted or applied. ## Match Each Question to a Method, Such as Interviews or Surveys Choose evidence sources according to the question rather than habit or tool availability. Existing evidence can establish context quickly, while new qualitative and quantitative research can explain motives, estimate prevalence, and test competing explanations. - Map each research question to a source, method, measure, and analysis approach. - Combine behavioral records, direct observation, interviews, surveys, and credible secondary sources as needed. - Use more than one source for decisions with material uncertainty or consequence. - State the strengths, weaknesses, and expected contribution of every method. ## Choose Who to Ask and Write Questions That Don't Lead Recruit participants who represent the relevant decision context, including important variations and overlooked groups. Design questions and tasks that minimize leading language, recall burden, social pressure, and ambiguity. - Define inclusion, exclusion, segmentation, and recruitment criteria. - Match sample size and diversity to the type of conclusion being drawn. - Pilot the instrument with representative participants and revise unclear items. - Keep response options balanced and separate observations from researcher interpretation. ## Collect Data the Same Way Each Time and Protect Participants Gather data through a repeatable process that preserves provenance and protects participants. Quality problems introduced during collection can create confident but false conclusions, so monitor execution while evidence is still being produced. - Train collectors and standardize instructions, definitions, and recording practices. - Capture source, date, context, consent status, and transformation history. - Check for missing values, duplicates, unusual patterns, and inconsistent coding. - Restrict access, remove unnecessary identifiers, and retain evidence for the approved period. ## Study Demand and Market Conditions Before Estimating Market Size Describe the forces that shape demand before estimating an opportunity. Consider category boundaries, customer alternatives, adoption barriers, regulation, technology, economics, seasonality, and the conditions that could make past patterns unreliable. - Define the market in terms of the need being addressed rather than a convenient label. - Distinguish observed demand, expressed interest, reachable demand, and paid adoption. - Compare current signals with historical context and plausible future changes. - State the assumptions behind any estimate of size, growth, or timing. ## Compare Competitors and Other Ways People Solve the Problem Study how people currently solve the problem, including direct rivals, substitutes, internal workarounds, and choosing to do nothing. Focus on verifiable customer value and market behavior rather than copying visible tactics or guessing private intent. - Compare target users, offers, pricing logic, channels, proof, strengths, and constraints. - Separate direct observation from inference and label confidence accordingly. - Look for underserved needs, switching friction, and advantages that are difficult to sustain. - Refresh time-sensitive findings before using them in a major decision. ## Build Customer Profiles from Real Evidence, Not Made-Up Details Create segments or personas from meaningful differences in needs, behaviors, constraints, and decision processes. Demographic labels alone rarely explain choice, and invented detail can make a profile memorable while reducing its validity. - Link every important attribute to observed evidence and a research source. - Describe triggers, desired outcomes, barriers, alternatives, information needs, and buying roles. - Distinguish the user, buyer, approver, influencer, and beneficiary when they differ. - Validate the model with people outside the original sample and revise weak assumptions. ## Compare Findings Across Sources and Show How Sure Each One Is Clean and organize the evidence before interpreting it, then examine both supporting and contradictory cases. Quantitative patterns show scale and relationships; qualitative evidence helps explain mechanisms, language, context, and exceptions. - Define metrics and coding rules before comparing groups or periods. - Triangulate findings across methods, sources, and researchers. - Test alternative explanations, selection effects, measurement error, and missing evidence. - Report confidence, range, sample limitations, and the boundary of each conclusion. ## Turn Findings into Ranked Actions with Owners and Review Dates Translate each material finding into a decision implication, action, test, or explicit reason to wait. Research creates value when it changes choices, while a recommendation without traceable evidence makes later learning difficult. - Connect findings to the original questions and distinguish answers from unresolved gaps. - Rank actions by expected value, confidence, effort, reversibility, and risk. - Define an owner, success measure, review date, and stop condition for each test. - Store the evidence trail and schedule a refresh when customers, competitors, or conditions may change.