Market Research – How to gather primary data

September 11th, 2009 by Lowell D'Souza Leave a reply »

market-research-primary-data-techniquesI was aware that brand building was an offshoot of advertising and wasn’t surprised to learn that market research was an offshoot of the advertising boom of the early sixties. Advertisers began to understand the value of getting to know their demographic better and this of course, was aimed at helping them sell more products.

Market Research, to me, is a two-way discovery process and is a key element of a marketing strategy. It involves the design of a plan to collect, analyze and report data/findings which answer a specific marketing situation that a company seeks answers for. This involves conversations, data analysis, strong processes, follow-through and continuous feedback.

The process is rather simple though the actual work involved may not be as simple. The steps involved are :

1. Define the marketing problem/situation and research objectives.

2. Develop the research plan.

3. Collect data.

4. Analyze data.

5. Present findings.

6. Make decisions.

market-truth-research-on-attitudes-towards-real-brands-on-second-lifeData sources for this exercise can be secondary data, primary data or both. Secondary data is data that was collected for another exercise and can possibly be sued for other research. Primary data is fresh data gathered for a specific purpose from a direct source.

This blog entry will focus on techniques on how to gather primary data. You can collect primary data in 5 ways : Observation, focus groups, surveys, behavioral data and experimentation.

1. Observation :

Before Ethnography, there was simple ‘ observation’. Here the customers and the setting in which they make their purchases are observed thoroughly. For e.g. if someone wants to purchase office productivity software, researchers observe the process with which the buyer walks around the store aisles while trying to determine the product he wants. The conversations and queries that the buyer has with the store support are also recorded. This kind of research helps the organization with the shelf-space positioning of their products as well as given them an understanding of any shortcomings with their products.

2. Focus-group research :

A focus group is a group of six to ten people who are invited to participate in a discussion about a product, service, brand, company or any other marketing issue. The group is managed by a skilled moderator who is objective, knowledgeable about the issue discussed and is an expert on group dynamics. The moderator beings by asking broad questions and slowly move the line of questions to the marketing questions that the organization seeks answers to. The discussion is typically recorded on video and is subsequently analyzed to understand customer beliefs and attitudes. If you’re cagey about group-think in a face-to-face focus group and worried about high costs of running a focus group, then Fulcrum ( who took over CyberDialogue) can arrange for a online focus group session where you can look into a chat room and see the entire discussion take place.

3. Surveys:

Surveys are undertaken by companies to learn about their customers beliefs, preferences, product knowledge and satisfaction levels. Typically, a small sample size is chosen as representative of the population and a series of questions are presented to them with a choice of answers. Objective answers usually have a range of values. For subjective answers, a field is provided to capture comments.

4. Purchase behavior analysis:

This is pretty much data mining. Most firms have customer databases where they have information about customer purchases, coupon usage, catalog purchases, online purchases etc. The actual purchase vs. preference myth is addressed here as its possible to understand if a certain high purchase demographic is buying premium brands or if a low-purchase demographic is buying cheap brands only. Different aspects of consumer behavior are also obtained from this exercise though it’s always wise to support this kind of analysis with focus groups or surveys.

5. Experimental research:

This kind of research is about cause-and-effect and is the most scientific. Here selected groups of subjects are exposed to different treatments with external variables in a limited or controlled environment and responses are noted closely to see if they’re statistically significant. A good example of this is A/B testing. For e.g. if a travel site offers an excursion package forĀ  $100 on all trips to Ireland for one week and then offers the same package for $125. Given that the number of visitors to their website varied marginally week over week, it’s possible to gauge the success of the two excursion packages pricing.

I’ll be writing more on the different tools that researchers use to capture primary data in another post. Till then, have a great Labour Day’s weekend!


Leave a Reply

5 × = five