You need to know if there is demand for a new service or product that you are thinking of introducing. Perhaps you need to know how your product is faring among the competing brands. You may want to assess brand awareness after your advertising and promotional campaign.

You want to find out the level of customer satisfaction. Or you could have been meaning to discover why your potential customers are still not responding.

The answers to your questions lie with your target customers and it appears that you need only to ask them.

Your research agency should help you to define your problems or questions to enable proper research design. Suppose you want to run an advertising and promotional campaign for a product that has been in the market for a while. You want to find out the impact of your campaign.

Your problem may be specified as:

a. Assessing the level of awareness of the product versus the competing brands before the campaign (the "mind share");

b. Estimating the market share before the campaign;

c. Repeating (a) and (b) after the campaign; and

d. Determining brand loyalty and brand switching tendencies by conducting brand tracking studies subsequently.

Research Design

The above problem obviously requires accurate generalisations to be made from the sample to the total number of your actual and potential customers (the population). Since it is not feasible nor necessary to ask everyone (elements) in the population, the first step is sampling or selecting an adequate number of "spokespersons".

Determining the Sample Size

Your agency will need to: 

• Estimate the population or the total number of possible customers, and 

• Determine an adequate sample to draw from the population.

What is adequate depends on the number of total possible customers you have. For example, if your product is exclusive, a high proportion of your customers may have to be sampled given the relatively small population. However, if your clientele is large, then a lower proportion needs to be selected. Sample sizes often vary anywhere from the hundreds to several thousands.

Sampling from the Population

Suppose feedback from 500 customers has to be obtained. It is pointless for only the loyal customers to be studied. Their positive assessment of your product would not be an accurate picture of your entire customer base, that is, not representative.

Thus, sampling has to be random, that is, everyone in the population has an equal chance of being selected. But this does not mean a haphazard way of selection. Due to our preferences, sampling is naturally biased because some elements in the population will have a higher chance of being selected while others may have little or none.

Your research agency basically has to utilise sampling techniques that will minimise subjective choice. Sampling is so crucial that you could be asking the wrong people and the entire research effort would have been a waste of valuable resources.

Data Collection

The research agency should choose the most appropriate method of collecting data from the sample. There are two broad methods of data collection: by observation and by questionnaires.

Collecting Data by Observation
This method of collecting data involves human observers as well as electronic or mechanical devices to record information and interesting behavioural traits.

Observation may be overt, which is open, or covert, which is concealed:

• Overt Observation

Sampled persons may be recruited to collect information about their household expenditure patterns, television and radio listening preferences, response to stimuli like advertisements and so on. Electronic or mechanical devices may be installed in homes to collect data of interest.

Customers may be invited to a research agency's "labs" to have their physiological changes (for example, heart rate and electrical resistance of skin), eye movements, voice and brain wave patterns analysed in relation to various stimuli. Diaries might be kept or surveyors may perform audits in homes and retail outlets to record the required information.

• Covert Observation

Human observers as well as devices such as cameras and optical scanners might be used to study behaviour in a natural setting. Shopper behaviour (for example, making choices among competing brands on a shelf in a supermarket), purchases made at cash counters, human traffic and their response to promotional displays and service quality evaluators disguised as customers are some examples. Since covert observation is unknown to the observed, ethical considerations have to be observed.

Data Collection by Questionnaires

The three main methods of collecting information by questionnaires are face-to-face interviews, telephone interviews and mail surveys. These methods have different strengths and weaknesses which would determine their suitability for your needs.

• Face-to-face Interviews 
  The questions you need to ask your customers may require interviewers to pose them personally. For example, your agency may advise that the possible responses to a question be displayed to assess aided recall or that a sample of your product be shown to gauge recognition. Or the questionnaire may simply be too complex to be completed by respondents themselves or administered over the telephone.
  
  Personal interviews allow greater flexibility in questionnaire design and enable higher response rates but are costlier and more time consuming than telephone interviews and mail surveys.
  

• Telephone Interviews
  Sometimes, the information you require may be gathered without the need to meet respondents. Suppose potential customers are to be asked if they have heard your radio commercials. They can be probed on, say, commercials on eating places that they may have heard over a specific period.

Telephone interviews are less expensive than face-to-face interviews and require less time but are restricted to simpler research.

• Mail Surveys
  There could be instances when data required cannot be provided "off-the-cuff'. For example, respondents might be required to refer to company records that have to be collated from various departments. Such questionnaires can be mailed to respondents but they have to be well designed to facilitate self administration.

Mail surveys are even more cost effective than telephone surveys. Naturally, however, without interviewers to persuade respondents to cooperate, response rates are the lowest. Thus mail surveys typically require incentives such as gifts even when there is vested interest, for example, allowing respondents to tell you how to improve your services to them.

Data Analysis and Solution Formulation

The data collected have to be processed and analysed in order to form a coherent picture of what your customers are telling you. The information collected by observation or questionnaires need to be checked for completeness and consistency.

Validation has to be done for questionnaires administered personally or over the telephone. Often data have to be scanned or manually entered into database management software to facilitate statistical analysis.

In order to be able to summarise the information, identify patterns and relationships as well as forecast trends, the large amounts of data have to be processed by statistical software.

Researchers typically apply a wide variety of statistical methods. They range from descriptive statistics (for example, frequencies, percentages and averages) to tests for association or difference between comparison groups to techniques for discriminating groups within the sample.

However, sound or powerful statistical techniques are only tools. A marketing orientation is needed to provide the basis for valid, meaningful and practical solutions and recommendations.

Presentation of Key Findings and Recommendations

While all research findings and recommendations can be found in the report, the verbal presentation to marketers and management is just as important since "on-the-spot" decisions often have to be made in relation to the findings.

Management may instruct that feedback from customers be incorporated to improve the company's services or that a new product line be introduced to cater to unmet needs as revealed by research. Perhaps future advertising budgets might be tightened to utilise only those media through which most of the targeted audience are reached.

Valid, accurate and timely quantitative research is essential to make vital marketing decisions concerning your entire customer base or clientele. With effective marketing research, marketers will be better able to increase market share, revenue and so on to achieve their marketing objectives.

Back To Top >>