Blog Marketing
Market research: how to do it with an example

Market research brings together information about customers, alternatives and market conditions so you can make a specific decision. It can help you decide what to change in an offer or whether to test it on a small scale. A few interviews do not estimate how many people would buy.
Start with the decision you need to make
Write the decision in one sentence. For example: “Should we test breakfast boxes for nearby offices, or change the offer?” Then define who might buy, where they are, when they would use it and over what period. Without those boundaries, a number from a broad category may look relevant even when it does not describe your possible customers.
The U.S. Small Business Administration’s guide recommends checking questions such as demand, market size, location, saturation and price. It also distinguishes existing sources from direct research with consumers. Use that list to frame questions; any data you use should match the location and category you are studying.
On mobile, scroll the table horizontally to read every column.
| Question | Evidence to gather | What it helps you decide | Limit |
|---|---|---|---|
| Who might buy, and for what occasion? | A potential customer’s situation and behavior, recorded with date and context. | Who to include in a test and which need to describe. | One conversation does not represent a whole market. |
| What alternatives do they use now? | Products, suppliers or workarounds people mention or you can document. | What would need to differ for them to consider your offer. | A list does not show that they will switch. |
| Which buying conditions matter? | Frequency, format, price, place, timing and relevant constraints. | What to include, charge or offer. | A hypothetical answer is not a purchase. |
| What can you check first? | Specific requests, reservations or a pilot with clear terms. | Whether to adjust the idea or take a limited next step. | A small test does not estimate market size by itself. |
Fictional example: breakfast boxes for offices
A bakery is considering breakfast boxes for morning meetings at offices within three kilometers of its shop. It wants to decide over the next two weeks whether to prepare a pilot. The bakery, responses and numbers are fictional.
First, define the question: is there a buying occasion the bakery can serve with its current hours and capacity? Before showing a box, the researcher talks with eight office coordinators about the last time they arranged breakfast: what they bought, for how many people, how much notice they had and what made ordering difficult. Eight is only the size of this teaching example, not a recommended sample.
Five people recall improvising breakfast for a meeting, and two agree to review a limited pilot offer. That supports one next step: confirm the contents, price, date and delivery with those two people, then offer a pilot only if the bakery can meet those terms. It does not show that five out of eight offices share the need or that either person will buy.
Four questions for those conversations
Think of the last meeting you ordered breakfast for. What did you buy, and for how many people? How did you choose the supplier, and what other options did you consider? How far ahead did you order, and how did the delivery go? What was hardest to arrange, and how did you deal with it?
If the problem is getting an order before 9 a.m. and the bakery cannot deliver then, changing the box design will not solve it. If offices need different quantities, the bakery should review pack sizes before printing a fixed catalog. Answers need to be able to change the offer.
Turn early evidence into a decision
- Review what already exists. Use business records and public data only when their date, definition, location and category fit the question. Secondary sources can describe the wider setting; they may not cover your specific offer.
- Ask about recent behavior. Ask the person to describe their last purchase or how they handled the task. Ask about alternatives, frequency, price and obstacles. Listen before describing your solution so you do not lead the response.
- Show a specific offer. Once you understand the situation, explain the format, price, timing and delivery. Ask what would prevent a trial. Record “sounds interesting” separately from a concrete action such as requesting a quote or booking.
- Agree on how you will use the result. Before testing, decide what evidence would lead you to adjust the offer, run a pilot or stop. Record conditions that could change the choice, such as capacity, cost or timing.
For this pilot, the bakery sets a condition in advance: it needs at least one request with a quantity, date and delivery time it can serve at a price covering the estimated cost of that order. The two people who agreed to review the offer have not yet met that condition. The bakery needs to quote and confirm those details before producing. This is the case’s criterion; each business needs its own, based on capacity and costs.
When you do not know the market yet
Do not present exploratory conversations as market percentages. If you need to estimate how many buyers exist, find a source that measures the right category, location and period, and explain what it includes and leaves out. If no comparable total is available, say that this part is still unknown. An estimate is useful only when you show its source and assumptions.
Market segmentation helps decide whether groups with different needs call for different offers or messages. The buyer persona guide develops research into a decision-maker for social content; it does not replace market sizing.
The SBA provides methodological guidance. Data for your category and location need to come from sources that actually cover them.
The cover is an AI-generated illustration.
Plan your next post in HeyMark.
Keep the idea, review the draft with your team, and see how it performed in the accounts you connected.