MKTG 203 Chap.4 Marketing Research for Decisions
Marketing Research for Decisions
Marketing research reduces uncertainty rather than guaranteeing the right answer. This chapter follows the five-step approach from problem definition to action, compares primary and secondary data, chooses qualitative, survey, observational and experimental methods by fit, and keeps conclusions within the limits of the design.
What this chapter covers
- 01
Define the decision problem first
- 02
Plan evidence before collection
- 03
Secondary data are fast but may not fit
- 04
Primary methods answer current questions
- 05
Findings must lead to bounded action
Worked example · free
Plan research for a refill station
- +2Define launch, redesign or stop as the decision.
- +2Use secondary category and competitor evidence first.
- +2Observe prototype use and explore concerns through interviews.
- +2Estimate prevalence only after themes are understood.
Key terms
- Marketing research
- A structured process for developing information that supports marketing decisions.
- Secondary data
- Information already collected for another purpose.
- Primary data
- Information collected for the current research question.
- Focus group
- A guided discussion with a small group used to explore perceptions and language.
Marketing Research for Decisions FAQ
Why does research start with problem definition?
Every later choice inherits the first definition. A vague question can produce interesting data that do not change a decision. State the action being considered, the uncertainty that separates the options and the evidence that could change the choice. Method selection then becomes a consequence of the problem rather than a habit.
When should I use secondary data?
Use existing evidence early to understand market size, trends, competitors and context at lower cost and speed. Check whether definitions, date, geography and population fit the current decision. Secondary data can narrow the gap but may not answer a business-specific question, which is when focused primary collection becomes useful.
Does a large sample guarantee a good study?
No. A large sample can estimate the wrong quantity precisely. Check whether the people reached represent the target, whether the situation resembles actual use and whether the measure captures behaviour or only stated intention. Sample size cannot repair a biased question, poor frame or irrelevant outcome.
What should come before choosing a survey or interview?
State the managerial decision and the uncertainty that blocks it. Only then choose evidence that can separate competing explanations. Beginning with a method can produce detailed data that do not change action. The final plan should name the result that would support launch, redesign, further testing or stopping.
Exam move
Memorise the five steps as a decision loop, then practise starting from a managerial choice rather than a method. For each case, write the decision, competing explanations and evidence needed to separate them. Build a comparison sheet for secondary data, observation, depth interviews, focus groups, surveys and experiments using fit, speed, depth, breadth, control and realism.
Challenge every conclusion with the design limitation: themes are not prevalence, reported intention is not purchase and one pilot is not every context. Finish each plan with a clear action rule stating what result would lead to launch, redesign, further testing or stop. That rule keeps research connected to strategy. Use a three-layer planning sheet.
At the top, state the managerial decision in a form that names the alternatives and timing. In the middle, list the information gaps that could change that decision. At the bottom, match each gap to the least demanding credible source. This order prevents a familiar method from becoming the research objective. For secondary data, record the original purpose, date, definitions and population before using the result.
For primary work, separate exploratory questions from measurement questions and causal questions. Interviews can reveal language and mechanisms, surveys can estimate reported patterns within a frame, observation can show behaviour in context and experiments can strengthen a comparison when treatment and outcome are defined. None removes the need to check sampling and implementation.
Practise repairing loaded, double-barrelled and hypothetical questions, then explain what bias each repair reduces. When reviewing results, write one supported conclusion, one tempting overclaim and one limitation. Repeat the exercise with a different sampling frame so you see how generalisation depends on who could be observed.
Finish with an action threshold set before the findings arrive, such as the level of repeat behaviour needed to extend a pilot. If the evidence misses the threshold, specify whether the correct move is redesign, further research or stop. A short executive paragraph should contain the decision, method fit, central finding, limitation and proportionate action.
That structure keeps the result useful without pretending that uncertainty has disappeared.
Working through Marketing Research for Decisions in MKTG 203? Sia is AskSia’s AI Marketing tutor — ask any MKTG 203 Marketing Research for Decisions question and get a clear, step-by-step explanation grounded in how MKTG 203 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
