MARKETING RESEARCH: AN OVERVIEW
Sensing the market is the most important factor for survival and growth of any enterprise. In a fast changing global scenario, this cannot be left to the instinct of the entrepreneur or manager. Very often in small enterprises, decisions are based on a salesman’s or sub-ordinates reports or second hand information. Unless the source of the information is authentic the decision could be wrong. Many decision makers fail to distinguish between information and knowledge. General information gathered may not relate to the problem and marketing opportunities may be overlooked. While gathering market intelligence, the manager must not just listen to the customer, but also empathize with him. This information would be available from two sources; from the internal records containing results and the market intelligence system relating to happenings in the market.
Generally a manager views the environment in four ways:
- Undirected viewing: This is information gathered from newspapers, trade journals and books read with no particular purpose in mind.
- Conditioned viewing: When a manger reads a sales report or anything similar on a specific area but which involves no active search, this called conditioned viewing.
- Informal search: When a decision maker interacts with retailers or agents personally on some specific topic he or she engaged in Informal search. Such an exercise would be rather limited in scope and unstructured in activity.
- Formal search: However when a manager goes about gathering information adopting a well thought plan and using a particular methodology, he or she is conducting a formal search, which is called Market Research.
THE MARKETING RESEARCH PROCESS
There are three types of research, which could be conducted. They are:
- Exploratory research: This is usually done to get preliminary data on the real nature of some problem
- Descriptive research; To describe certain magnitudes, like how many might avail of a conference hall facility in a hotel.
- Causal: This is intended to test the cause-effect relationship between factors. For example if service charge in a hotel is reduced from $500 to $400 would it result in an increase of 20% in those availing of business center services?
Distinction between Market research and Marketing research:
There is a common tendency to use these two words synonymously , but the two are distinctly different.
Market research is concerned with:
- Estimating consumer demand
- Identifying existing and potential demand
- Nature and size of the market
- Consumer reaction to existing product
- Consumer’s possible reaction to new products
- Degree of competitiveness
Marketing research on the other is more comprehensive. Market research is a part of Marketing research but it also encompasses the following:
- Product research
- Sales and marketing research
- Consumer research
Product research would comprise of:
- Product studies
- Testing of existing products
- Packaging research with emphasis on design and physical characteristics
Sales and Market studies would focus upon:
- Determination of market characteristics
- Market share analysis
- Market potential
- Sales analysis
- Estimating and establishing sales quotas and territories
- Test markets
- Store audit
Consumer research would b concerned with
· Identifying customers and their characteristics
· Customer buying habits and preferences
· Customer motivation
With the growing importance of Market Information, other kinds of research too are conducted by most organizations. They are:
Advertising research which studies
- Advertising effectiveness
- Copy research
- Media research
Business and Economic corporate research which tracks:
- Business trends
- Short range forecasting
- Long range forecasting
- Pricing studies
- Plant and warehouse location studies
- Operation studies
These days owing to social impact of an organization corporate responsibility research is also conducted to assess the ecological impact of the companies’ activities.
Every marketing plan involves five basic steps as shown below:
- Defining the problem and objective of the research
- Prepare a Research plan
- Gather Information
- Analyze the information
- Presenting the Finding of the study
DEFINING THE PROBLEM.
The success of any kind of research depends on defining the problem. In marketing research if the problem is not well defined, not only will the gathered information be useless but also prove costly .The more precise and specific it is, so much the better. The problem should not be defined too broadly, nor should it be very narrow.
Let us take an example from the hospitality industry. This service industry is not only very competitive but also functions in a complex socio-cultural milieu in different parts of the world. Very few hotels chains dare to function in some of the conflict zones like Afghanistan and Rwanda. Some like Serena hotels provide elaborate security to their clients’ security. Let us assume that XYZ Hotel intends to provide personalized security to some of their high profile customer and plans to conduct a marketing survey.
Broad definition of the problem: Find out the security needs of those checking into the hotel.
Narrow definition of the problem would be: If $700 is charged extra would the hotel breakeven on the new service offered.
The hotel would definitely be interested in knowing answers for some of these questions:
- What are the main reasons for availing personalized security
- What kind of customers would need it?
- How many would probably avail of this services and what would be the best price to charge
- How many extra customers might check in on account of this new extra service.
- What would be the relative importance of this service in relation to other services like food, entertainment, transportation etc.
THE RESEARCH PLAN
Facts can be collected in two different ways. Either on the basis of data collected by the investigator or on the basis of data readily available. The first is called Primary data and the other, secondary data. The investigator who gathers Primary data is engaged in Primary research and those who depend on Secondary data is engaged in Secondary research. The first thing an investigator has to decide is whether he or she should depend on Primary or Secondary data. This is because Primary research is a lot more costly than Secondary research. If the cost of Primary research is lot more than the data collected in terms of value, it is not worthwhile to proceed in those lines. Primary data is needed when the secondary data is either outdated or the investigator needs to do exploratory research. Success in primary research however depends upon two important factors. One is the sample selected for conducting the survey and the other is the questionnaire that has been prepared for eliciting information.
SOURCES OF DATA
The easiest source of getting facts is from Secondary data, which consists of information that has already been collected by others for other purposes. Secondary data would be available from both external and internal sources. Some of the most common sources of external data are:
- Official publications of federal (central), state and local governments.
- Official publications of world bodies like IMF, WORLD BANK, ILO, WTO etc (examples are National Accounts Statistics, Statistical year book, Civil Affairs Handbook are just a few noteworthy ones)
- Reports and publications of trade associations, chambers of commerce, stock exchanges, banks etc.
- Newspapers (both general and business), monographs, books, technical journals, trade journals etc
- Reports submitted by commissions/committees, economists and research scholars.
Internal sources of data are:
- Company profit and loss statement,
- Balance sheets
- Annual reports
- Sales reports
- Ad hoc research reports
As mentioned earlier though secondary data is low cost and quickly available, it may not meet the requirement of the investigator in which case he or she would have to do primary research.
Primary data can be elicited in three ways, they are:
In this exploratory way of gathering data, emphasis is on the basis of observing the relevant actions of the subjects. The most common and simplest method is Personal observation. Here the investigator observes the actions performed by the subject (respondent) in the actual setting. So if the researcher is interested in knowing the respondent’s behavior in a supermarket the investigator adopts a non-intrusive method of observing the actions of the subject. If the researcher desires a greater level of accuracy, mechanical observation may be used. Here the researcher uses devices like eye-tracking monitors, pupilometers etc to probe the subjects response to commercials on TV and measure their level of interest. Another method used is conducting audits like Pantry audit. In this case the investigator takes stock of the brands, quantities and pack sizes in the home of the consumer while conducting channel studies in marketing. Another method used in observational studies is the use of Trace analysis. This is done on the basis of finding evidences of actual physical traces of use. For example by determining the number of fingerprints in a page of a magazine, it is possible to gauge the readership of various advertisements.
Focus group research
A second source of gathering data is by interviewing about 6 to 12 people who are either experts, users of the product or in some way related to the area of activity. Conducted in a informal setting this helps in providing a glimpse into consumer perceptions, their attitudes and levels of satisfaction.
This kind of research is usually done to establish a causal relationship between two apparently unrelated factors. A common example cited is the linkage of the use of tobacco with the incidence of cancer. Many who propagate the ban on smoking highlight the probability of the users of tobacco having a greater chance being afflicted with cancer. While trying to find out any causal relationship, the process involves first in selecting matched groups of subjects and ensuring that extraneous variables are controlled. When doing experimental research the investigator has to check whether the observed response differences are statistically significant.
In this method the investigator gathers the required information directly from the respondent. It may be to estimate the market size, audience habits, viewing patterns or any thing similar from a statistical population. When a count of every unit of a statistical population is carried out it is called Census. As this can be difficult and costly, it would be better to conduct the survey amongst a group of individuals who are representative of the whole population. The technique of using a mathematically selected portion of a statistical population to describe or evaluate it is called sampling. While doing this the investigator should be clear about who is to be surveyed (sampling unit), how many to be surveyed (sample size) and the geographic area where it is to be conducted (sample area). Success of the survey depends upon the absence of bias which is bound to creep in and is called sampling error. This is the deviation between the observed characteristics of a sample and those of the population, which it represents. Such errors may happen inadvertently, and need not be deliberate.
To avoid such bias the investigator would adopt a sampling procedure. The basic concern here is how should the respondents be chosen? Two very popular methods are probability sample and non-probability sample.
Probability sample is a statistical sample wherein all population members in the universe being measured have an equal chance to be included. The different types of probability samplings are:
- Quasi random sampling
- Stratified random sampling
- Cluster sampling
- Multi-stage sampling.
Non-probability sample is not as scientific as a probability sample. In this case the sample is selected on non-random basis usually based on interviewee availability and is subject to the bias of the investigator. Non-probability sampling is also known as convenience sampling, Quota sampling, judgment or purposive sampling.
Personal contact: Though this is probably ideal, it can be costly and the interviewer needs to be trained to do it professionally. He or she needs to be trained not only in following the dos and don’ts of conducting an interview, but also to be a good ‘Listener’
Telephone surveys: Telephone surveys can be depended upon only if a majority of the respondents have telephone connection and their co-operation in answering to a person who they cannot see. The interviewer must also be trained to not only present the questions clearly but also record it accurately. A new method which has been developed to reduce the data entry time is by making use of Computer assisted telephone interviewing (CATI) in which the interviewer reads out the questions displayed on the computer monitor and enters the response given immediately into the computer using the keyboard. A later modification is Direct Computer interviewing (DCI) where the computer directly interacts with the computer. In both these cases however it is effective for closed ended questions only.
Mail survey though cost effective may not elicit the same kind of response as personal contact or even telephone surveys. A new modification to this is however is the use of e-mails to which is comparatively better in terms of response.
In all the above cases four kinds of problems may arise:
- Non-availability of respondent
- Non-cooperation of the respondent
- Biased/dishonest respondent
- Biased/dishonest interviewer.
After the right sampling, the next most important research tool is the questionnaire. Only relevant questions should be asked and framing of the question is very important. As questions can be asked in many ways, they have to be not only carefully developed but also pre-tested before it is applied. In a questionnaire form, the form, wording and sequence are very important.
There are two types of questions that are normally used--- close ended and open ended.
Close-ended questions are easy to quantify and can be subjected to statistical calculations.
A close-ended question could be dichotomous, which means offering two answer choices for a question. For example:
- Did you personally phone to book your room in this hotel? Yes No.
Or offering multiple choices
- How often do you check into this hotel?
Once a year
Once in six months
Once a month
Once a fortnight
In qualitative research, the investigator may use rating scales to measure a respondent’s subjective opinion like like/dislike, agreement/disagreement. The interviewee would be asked to mark their choice on the basis of the following options provided:
- Poor (1) Fair (2) Good(3) Excellent(4)
Open-ended questions are used to record the individual respondent’s opinion on a particular issue .It may be in the form of word association, like asking ‘what is the first brand that comes to your mind when you think of tooth paste’ or it may be sentence completion like ‘when I want to check into a hotel it would be…’
Once the required data is collected; then starts the laborious process of analyzing the information gathered. Once laboriously performed just using a calculator and statistical formulas, there are now statistical software, which reduce this burden. One such software is SPSS 8.0 (Statistical package for Social Sciences version 8.0). Apart from basic statistical, survey analysis, quality control and forecasting the new version has enhanced graphics capability, a simplified e-mail distribution of report and options for web publishing.
Presentation of Report
Unlike academic report, business reports are functional and so differ essentially in the format. Depending on the nature of the problem and its importance, it may be either a long report or a short one. The basic format of a long report would be as shown below:
Letter of authorization
Table of Contents
Table of Illustrations
Objective of the study
Scope and Limitations of the study
Findings and conclusion