The Marketing Research Process: Six Steps

Most marketing decisions involve a bit of guesswork. Will customers like the new product? Is the price right? Why are sales dropping? Marketing research is how we replace some of that guesswork with evidence. But research only helps if it is done in a sensible order.

That order is called the marketing research process. Different textbooks split it up in slightly different ways. Kotler and Armstrong’s Principles of Marketing uses four broad steps, OpenStax uses seven, and Naresh Malhotra’s well-known research text uses six. The ideas underneath are the same. In this article we’ll use the six-step version, because it shows each stage clearly without getting too fiddly.

A Running Example: Tom’s Café and the New Lunch Menu

Imagine Tom runs a small café near a university campus. Mornings are busy with coffee orders, but lunch is quiet. Tom is thinking about adding a new lunch menu of toasted sandwiches and salads, and he wants to know whether it’s worth the cost of new equipment, extra staff hours and wasted food while he finds his feet.

Tom could just go ahead and try it. Plenty of small business owners do. But a few weeks of simple research could save him a lot of money if the idea is wrong. So let’s walk Tom through the six steps.

The Six Steps of the Marketing Research Process

Step 1: Define the Problem

This is the most important step, and it’s the one that gets rushed most often. If we define the wrong problem, everything that follows is wasted, no matter how well it’s done.

For Tom, the obvious problem looks like “Will people buy my new lunch menu?” But is that really the problem? The underlying issue is that lunch sales are low. Maybe customers don’t know he serves lunch. Maybe students bring food from home. Maybe the café down the road is cheaper. A new menu only fixes one of these.

So a better problem statement might be: “Why are lunch sales low, and would a new menu increase them enough to cover its costs?”

Defining the problem usually means talking to the decision maker (here, Tom himself), looking at existing information such as sales records, and sometimes doing a little informal research, like chatting to a few regular customers.

Step 2: Develop an Approach to the Problem

Next, we turn the problem into specific research objectives and questions. What exactly do we need to find out?

Tom’s objectives might be:

  • Find out where his morning customers currently buy lunch, and why.
  • Measure interest in toasted sandwiches and salads.
  • Find out what price customers would expect to pay.
  • Estimate how many extra lunches he might sell each day.

At this stage we might also form a hypothesis, which is simply an educated guess we can test. For example: “Students would buy lunch here if it was priced under $12.” Having a clear hypothesis keeps the research focused.

Step 3: Formulate the Research Design

The research design is the blueprint for the whole project. It answers questions like: What data do we need? Where will it come from? How will we collect it? Who will we ask, and how many?

The first choice is between secondary data (information that already exists) and primary data (information we collect ourselves). We cover this in more detail in our article on primary vs secondary data in marketing research. Tom already has some secondary data in his till records, which show sales by time of day.

For primary data, Tom needs to choose a research approach. He could observe how many people walk past at lunchtime carrying takeaway food. He could run a short survey. He could even run a small experiment, offering a test menu for one week. He also has to decide whether he wants numbers, words, or both, which is the difference between qualitative and quantitative research.

Then there’s the sample. Tom can’t ask everyone, so who should he ask? If he only surveys his morning coffee regulars, he’ll miss the students who never come in at all. That could give him a very one-sided picture.

Let’s say Tom settles on a simple design: a short paper survey of 150 customers over two weeks, plus a count of passing foot traffic at lunchtime on three different days.

Step 4: Collect the Data (Fieldwork)

Now the research actually happens. This is often called fieldwork, because researchers go out “into the field” to gather the data.

For a large company, fieldwork might involve a research agency running thousands of online surveys or interviews. For Tom, it means handing surveys to customers with their coffee and having a staff member count passers-by from 12 to 2pm.

What can go wrong here? Quite a lot. Staff might forget to hand out surveys on busy days. Customers might rush through the questions. A long questionnaire can lead to survey fatigue, where people stop answering carefully. So Tom keeps the survey to five or six questions and gives his staff clear instructions.

Step 5: Prepare and Analyze the Data

Raw data is messy. Before we can analyze it, we need to check it. That means removing incomplete or obviously silly responses, coding the answers so they can be counted, and entering everything into a spreadsheet.

Then comes the analysis. For Tom, this could be fairly simple. Suppose 150 surveys come back and 130 are usable. Of those, 70 customers say they’d be likely to buy lunch at the café, and most of them say they’d expect to pay between $10 and $13. His foot traffic count shows a steady stream of students carrying lunch from a food court five minutes away.

Does this mean Tom will sell 70 extra lunches a day? No. People often say they’ll buy something and then don’t. So Tom needs to read his results with a bit of caution. A sensible analyst treats stated interest as an upper limit, not a forecast.

Step 6: Prepare and Present the Report

The final step is turning the analysis into something a decision maker can actually use. In a company, that’s usually a written report and a presentation with clear charts and a short list of recommendations.

A good report doesn’t just dump numbers on the reader. It goes back to the original problem and answers it. For Tom, the answer might be: “There is real interest in lunch, mainly at the $10 to $12 price point, but customers also mentioned slow service at busy times. A small menu trial is worth running, as long as orders can be served quickly.”

Notice that the research has also turned up something Tom didn’t ask about: speed of service. That’s common. Good research often reveals a second problem worth knowing about.

Why Does the Order Matter?

Why not just jump straight to a survey? Because the survey questions depend on the objectives, and the objectives depend on the problem. If Tom had skipped Step 1, he might have spent two weeks asking about sandwich fillings when the real issue was that students didn’t know the café served lunch at all.

The process also isn’t always a straight line. Sometimes the analysis in Step 5 shows that the problem was defined poorly, and we have to loop back. That’s not a failure. It’s research doing its job.

Thinking Like a Marketing Manager

If you are the manager paying for research, here are a few practical questions to keep in mind:

  • Is the problem clearly defined before any data is collected?
  • Is the cost of the research sensible compared with the cost of getting the decision wrong?
  • Does the sample include the people we actually want to reach?
  • Are we reading the results honestly, or looking for evidence that supports what we already wanted to do?

That last one is a real risk. Tom loves the idea of a lunch menu, so he’ll be tempted to see the survey results in the best possible light. Part of good research is being willing to hear “no.”

So here’s a quick test. A clothing brand notices online sales are falling and immediately commissions a survey about new colors for next season. Which step have they skipped? They’ve jumped past defining the problem. Falling sales could come from price, delivery times, a competitor or the website itself, and color may have nothing to do with it.

Sources

Wikipedia: “Marketing research process”

OpenStax, Principles of Marketing: “6.3 Steps in a Successful Marketing Research Plan”

Scroll to Top