Market research can be categorized along two axes: Qualitative vs. Quantitative, and Primary vs. Secondary.

Qualitative research cannot be (or should not be) analyzed statistically. You are looking for anecdotes, scoping information, insights from individuals. This often includes discussions, focus groups, interviews, etc., where you are attempting to gain expertise and familiarity with a topic and not trying to rank various items or get a definitive answer.

In general, qualitative research happens before quantitative research because it helps you define the "waterfront" of the topic you are exploring. Thorough qualitative research helps define all of the critical topics relevant to a research project so that you don't leave out anything important in the quantitative phase. This output becomes a critical input to the quantitative research.

Quantitative research involves gathering data or responses in a consistent manner so that you get data that can be analyzed statistically. The most common example of quantitative research is the survey, where researchers contact a large pool of individuals by phone, email or in person and ask them a specific set of questions in a consistent fashion.

How you choose your respondents matters. When people self-select to take a survey (for example, CNN's homepage surveys or American Idol voting), you have introduced bias into the selection process; respondents are likely to have a vested interest or a preference in the outcome. It's not a random group. For example, if you set up a survey on Yahoo's home page that asks, "Should the US annex Canada?" the results will come primarily from people who have a strong interest or opinion in the topic. Thus the results will not be objective and do not represent the broader population.

Primary research consists of when you interact directly with consumers, customers, end users, etc. and collect original information and data that did not previously exist. So if you are directly talking to calling or emailing people and asking them questions to get their views, you are conducting primary research.

Secondary research is information that someone else collected and developed and that you are reviewing to see if it has implications for you and your business. This often includes reviewing articles, getting "Google alert" notifications on various companies, etc.

Uses of Research by Category

Qualitative Primary Research (top left box)

Focus groups are typically conducted in a specialized facility with a professional moderator leading a group discussion. The party sponsoring the research usually has some interested individuals behind a one-way mirror to observe and occasionally send in questions to the moderator. We often use focus groups to test survey questions we intend to use in the later quantitative phase.

The benefit of this is that the focus group participants often help you think of related, additional questions and, even more importantly, they will ensure that the questions you ask will be clearly understood by survey respondents.

Interviews are exactly what you might think. Researchers either telephone respondents or find them in person (e.g. "mall intercepts") and ask them for detailed answers to open ended questions.

"Mystery shopping" involves hiring a third party to buy from your company and from competitors to compare the experience from end to end. This is invaluable and often very painful to see how you stack up to your most hated rivals.

"Competitive digging" involves hiring a third-party to call your competitors and ask for information. There are tight ethical standards about what the "digger" is able to say and they should never tell lies. Amazingly, a great digger can frequently find a knowledgeable and indiscreet contact who works for a competitor, has access to lots of information and is willing to share it simply because a nice-sounding person makes them feel important by wanting to hear their opinions.

Qualitative Secondary Research (bottom left box)

For many years, professional market researchers relied on "clipping services" like Bacon's, which employed a large group of people to wade through every publication they could find — from national and international magazines down to local weekly newspapers. They would find every company mentioned and send "clippings" of these articles to paid subscribers.

Today, this service has largely been replaced by Google Alerts. Google Alerts are free; you simply enter a list of terms into the Google alerts database and the search engine emails every link to you that it finds that match your search terms.

Anytime you review research conducted by others, you are doing "qualitative secondary research" because you are not talking directly to customers and you are not running statistically analysis on the underlying data — you are simply reviewing someone else's conclusions.

Quantitative Primary Research (top right box)

In our experience, this is the most powerful and useful research of all — and also the most expensive, the most prone to error and the most reliant on following demonstrated best practices for accurate output.

The most common form of quantitative research is the survey. This is where you ask a randomly-selected audience a set of questions in a specific and consistent manner. The many facets of good survey design, administration and analysis are beyond the scope of this document but it takes a real expert to do it all well.

Surveys can take different forms, from "CSAT" (customer satisfaction) to "conjoint analysis," a methodology consumer products companies use to determine the right mix of features and price for any given product.

Quantitative Primary Research (bottom right box)

Quantitative primary research involves either conducting your own analysis of someone else's primary data or doing quantitative analysis of secondary data. Sound confusing? It's not. In the first case, if you take, for example, the US Census database and run statistical analysis against it, you may come across important marketplace findings that no one else knows. Similarly, if you count how many times your company is mentioned in positive and negative contexts compared to your competitors, then you may have some understanding of the relative brand equities of the companies.

• Categories of Research
• Market Analysis and Measurement
• Observations on Research