Market Research
Structured market and opportunity research.
Research can itself become a product.
Evidence-backed reports, market intelligence and decision-support packages with traceable sources.
MOJPLAYWIN / CURATED
Industry Briefs
Focused commercial intelligence packages.
Localization Research
Language and market adaptation research.
From a research question to a decision-ready report
Begin with a decision that someone actually needs to make. Frame the question, target population, outcome measures and time horizon before selecting a method. For market research, record data collection dates and sampling bias; for a product experiment, decide what change counts as meaningful before looking at results. A source list is not sufficient without an assessment of reliability and relevance.
Separate observations from causal claims. A rising conversion rate after a redesign might reflect a different traffic mix rather than the design itself. Use comparative groups when feasible and state limitations when they are not. A useful report presents the strongest evidence, competing explanations, downside risks and a reversible next action. Preserve source links, data transformations and version history so another researcher can reproduce the conclusion.
Apply this in three steps
- State the decision, assumptions and measurable outcome.
- Collect at least two independent forms of relevant evidence where possible.
- Publish a concise conclusion with confidence level, limitations and the next test.
Key takeaway: Good research changes a decision—or shows why no decision is justified yet.
Explore more free field notes →Estimating market size without inventing demand
Market size is a hypothesis built from definitions and evidence, not a large number copied from a presentation. Differentiate the total addressable market, the portion a company could realistically serve and the share it might acquire under a credible launch plan. Count actual potential buyers rather than every person who might be interested.
Working method
Use a bottom-up model: eligible customer organizations multiplied by a plausible purchase frequency and price per order. State exclusions, geographic coverage, data dates and how the sample was selected. Compare the bottom-up answer with independent sector data to expose implausible assumptions.
Worked example
If five hundred eligible firms each need two small research projects per year at a hypothetical three hundred dollars per project, the arithmetic opportunity is three hundred thousand dollars annually. This does not imply contracts exist, that firms will buy or that the company can capture the market. Survey buyers and measure conversion before forecasting revenue.
Common mistake
Unverified forecasts must never be reported as audited sales or signed demand.
Action checklist
- Write the buyer definition.
- Document each model assumption and source.
- Test willingness to pay with a small pilot.
Building a reproducible source and evidence register
A research report is more useful when another reader can reproduce its main claims. Keep an evidence register connecting each finding to the original source, access date, method, reliability assessment and any transformation made to the data.
Working method
Record whether a statement came from a primary document, a measured observation, an interview, a statistical inference or a model assumption. Preserve original measurement units and sample sizes. For calculated figures, include formulas and a short explanation of missing values and exclusions.
Worked example
Imagine a report comparing agricultural export opportunities across two regions. A dated customs dataset may measure historical trade volume, while interviews indicate current distributor interest. The first is observed market activity; the second is qualitative evidence. Combining them into one figure without noting the difference would overstate certainty.
Common mistake
When important evidence conflicts, describe the competing interpretations rather than deleting the inconvenient source.
Action checklist
- Number the claims that influence decisions.
- Link each claim to verifiable supporting evidence.
- Keep a versioned methods and limitations appendix.