# Market research basics

The method follows the decision: published studies, your own survey, qualitative work, or the data the business already generates.

In January 2013 the post answered a Quora question, "How do you conduct market research?", by quoting Cate Riegner. The four methods below are hers. One spelling error in the original ("stuides") is corrected. The sample warning is updated.

## Four methods

Think about which method fits the stage of the business, the product, and the marketing decision.

**Syndicated and secondary research.** This is work already done on a market: size, penetration, trend. Look at trade bodies and published studies. The hard part is finding one that matches your question. Many of the useful ones charge a fee.

**Your own quantitative survey.** Use this when you need numbers on a specific subject: behaviour, attitudes, or a product concept. The sample has to represent the market you claim to describe. The questionnaire has to avoid leading questions and use scales people can answer. The tables have to be processed and reported without bending them. Hosting tools such as SurveyMonkey make the form easy to field. They do not supply a representative sample, a sound questionnaire, or the analysis. Those are separate costs.

An online form filled in by whoever sees the link is a convenience sample. It is not a probability sample of Jamaican households or firms. If the decision needs that kind of estimate, the design has to start from a frame and a selection rule, whether the interview is face to face, by phone, or online.

**Qualitative work.** This covers user testing, ethnography, interviews, blind shopping, and focus groups, in person or online. Quantitative work gets more deference because people treat it as statistically firmer. A small qualitative sample can still be the right tool when you need to understand a decision, not estimate a percentage. Pick the approach that matches the question. The methods above do not answer the same question.

**Data the business already generates.** Traffic logs, sales records, and similar streams are often free. The work is collecting them and tying the finding to the question you actually have.

Riegner's full thread was on Quora. The link in the 2013 post no longer resolves to a stable page, so it is not repeated here.

## Updated October 2026

- "Blind-shopper stuides" is corrected to "blind shopping". The four-method argument is otherwise hers, as the 2013 post presented it.
- The sample warning is tightened. A self-selected online panel, including a hosted form, is not a probability sample. That limit was implied in 2013 and is stated outright now.
- No new market-size figure is added. The post did not contain one.


Source: https://balcostics.com/blog/market-research-basics/