A market research report fails in two common ways. It is a data dump: everything is included, and nothing has a point. Or it is a collection of opinions: there is a point, but no evidence behind it. This structure avoids both. It forces a question, a source plan, and a storyline, so the report ends up as an argument a client can follow and a recommendation they can defend. It is the same structure we use before any entry, expansion, or investment decision.
Step 1: Write the question before the research
A report answers a question, not a topic. "What is the market for X" is a topic. "Should we enter market X, and if so, with what positioning" is a question. Write the question at the top of the document in one sentence, and let it filter everything you collect.
The discipline matters because the question determines what is evidence and what is noise. A report built to answer a positioning question needs customer behaviour and competitive data. A report built to answer a go/no-go decision needs market sizing and entry costs. When a piece of data does not move the question, it does not belong, no matter how interesting it is. Keep the question visible while you work, and test every new finding against it.
Step 2: Plan the sources
Map each section to a source type before you start collecting. Registry and government data for market size, industry surveys for adoption rates, competitor pricing pages for price points, and interviews for customer behaviour. A source plan written in advance does two things: it stops you from burning expensive interviews on facts you could have read, and it shows you where the evidence is thin before you have invested weeks in collection.
The quality of a report is the quality of its sources, and the honest note is that weak sources make a weak report no matter how clean the charts are. For every section, decide what would be good enough evidence: a government statistic, a verified interview, a competitor price list, an analyst estimate. Then match the section to the best available source and write down what you will use. If a section has no credible source available, that finding is a gap to flag rather than a blank to invent.
Step 3: Collect into one place, with provenance
As you gather, keep the source attached to every figure from the moment you record it. A number with a source and a date is traceable. A number without one is a liability. The figure you cannot trace back is the one that will get challenged in a review, and it is also exactly where fabricated data shows up. If it has no source, it does not belong in the report.
Set up a simple collection table up front with four columns: the figure, the source, the date collected, and the note on how it was calculated. This looks like overhead at the start, but it is what makes the writing phase fast. When you build the report, every number is already sitting next to its source and its calculation, so there is nothing to hunt for and nothing to misremember. Keep the raw material separate from the report itself, so the report stays a curated argument rather than a dump of everything you found.
Step 4: Build the storyline
Organize the report around the argument, not around the data. A typical structure runs: the opportunity, the segments, the customer behaviour, the competitive dynamics, and the recommendation. Each section should earn its place by moving the answer forward, and each one should set up the next.
Write a one paragraph outline before you write the report. If you cannot say in a paragraph what the report concludes and why, the structure is not ready. Then test each section against the question from step one: does this section change what a reader would decide? A section on industry history that does not affect the recommendation is decoration. Cut it, or shrink it to a footnote. The reader should finish the report knowing exactly what you are recommending and what evidence supports it, in that order.
Step 5: Label facts vs estimates
A fact has a source you can open: a published statistic, a verified interview, a documented price. An estimate has an assumption you can state: the calculation you used and the logic behind it. Both belong in the report. The crime is mixing them so the client cannot tell which is which.
Write the assumption next to every estimate, in the same cell or line as the number. "I assumed 40% of SMEs outsource this" tells the reader exactly how much weight to give the figure. A third decimal place with no stated assumption is worse than a round number with one, because it looks precise and is not. If you are unsure whether something is a fact or an estimate, label it as an estimate. The reader can always make a fact out of an estimate by verifying it, but they cannot trust a number that misrepresents its own certainty.
Step 6: Show the range and the sensitivity
End with the range and what would move it. Give the pessimistic, base, and optimistic numbers for the key figure, and state the single assumption that drives the difference between them. Clients trust a report that tells them the uncertainty honestly, and the recommendation is stronger for having survived the pessimistic scenario.
A worked example: your market size estimate for a niche software category comes out at $40 million. The base case assumes adoption by 30% of eligible firms. The optimistic case assumes 40%, the pessimistic case 20%, and the resulting range runs from roughly $27 million to $53 million. The client now knows exactly what would have to be true for each end of the range, and they can watch the right signal. That is more useful than a confident point estimate with no basis, and it survives the review meeting.
Common mistakes to avoid
- Skipping the question. Reports written without a question drift into a summary of everything found. The fix is a written question on page one.
- Collecting before planning sources. Weeks of research can produce a pile of data with no coverage of the sections that matter. Plan the sources first.
- Mixing facts and estimates silently. One unlabelled guess can undermine a whole report once it is spotted. Label every estimate.
- Presenting the data in the order it was found. The report must follow the argument, not the chronology of your research.
- Hiding the uncertainty. A single point estimate invites a challenge. A stated range with the driving assumption is far easier to defend.
Speed it up
The research and collection phase is the compressible part. Use the free market research tool to structure the question, sources, and storyline up front, then spend your effort on the evidence and the argument.