MARKETING RESEARCH

Marketing research
This is the means used by an organization to keep in touch with the needs, wants and
attitudes of those who purchase or could purchase the organization's products and
services.
Frequently, financial services organizations need to commission specific marketing
research studies in order to have adequate information to make decisions.
Such research can involve:
a) Product research to assist in:
• the design, development and testing of new products and services
• the improvement of existing products and services
• the forecasting of trends in customer preferences for specific products and
services.

b) Market research to assist in:

 the analysis of market segments, market size and the share of the market held
by the financial services organization and its competitors.
c) Customer research to assist in:
 the analysis of buying behaviour
 the analysis of customer perceptions, attitudes, reasons for purchase.
d) Promotion research to assist in:
 the testing and evaluation of promotional material.

Any of these types of research can be undertaken as part of a continuous research

project where changes in market share, consumer attitudes, etc can be tracked over a
period of time. Alternatively, an ad hoc study is undertaken on a one-off basis and
provides the organization with a snapshot of what the situation is at any point in time.
The research would normally be undertaken either by the organization's own
marketing/market research department or by an external marketing research agency.

There are advantages and disadvantages of using external agencies.

The advantages relate to the agency having specialist skills, facilities and resources
supported by the fact that their analysis of the findings can be undertaken with greater
objectivity as they are not concerned with the internal politics of the organization.

The disadvantages relate to risks of information leaks to competitors and the

possibility of valuable information being lost in the communication of findings
between the agency and the client company.

Research agencies can be categorised into the following types:

Syndicated service agencies
These agencies specialise in gathering continuous consumer and corporate
information which they sell in the form of standardised reports to any company
wishing to purchase them.

Ad hoc research agencies

These agencies are hired by a client to carry out one-off research projects for the sole
use of the client company. They participate with the client in designing the research
study and the final report becomes the client’s property. They can take one of the
following forms:
i. Full service agencies – able to do all types of market research in almost any
market
ii. Market specialist agencies – has specialist skills in carrying out research in
specific markets (ie the corporate market, the consumer market, amongst
children, in the financial sector or in certain geographic regions)
iii. Technique specialist agencies – has specialist skills in carrying out research
using specific techniques (ie telephone research, the use of street
interviewers, small in-depth studies).
Fieldwork agencies
These agencies provide interviewers either to other research agencies or to client
companies. They are only responsible for carrying out the interviewing, with all other
design and analysis functions being undertaken by the company which hires them.

THE MARKET RESEARCH PROCESS

Stages in the market research process

UNDERSTANDING THE BUSINESS PROBLEM OR OPPORTUNITY- DEFINITION OF RESEARCH

OBJECTIVES - RESEARCH DESIGN - DATA COLLECTION - DATA ANALYSIS & REPORT
PRESENTATION

Understanding the business problem or opportunity

It is important that the reason for undertaking the research is clearly understood,
before actually defining the research objectives. It is all too easy to undertake research
on areas that would be nice to know; however, if there is not a commercial reason for
undertaking the research, it is simply an expensive pastime. Defining the problem will
clearly identify if research is needed and what information is required to assist the
decision makers in tackling the business problem or opportunity.
The problem itself could be signaled by a number of events, such as a failure on the
part of the organization to attain its operating objectives, a drop in sales, an increase
in expenses, or a fall in profits. Whilst these are all negative examples, the stimulus to
market research could be something positive which the company wants to get to the
bottom of and do more of. Examples of this would include an increase in sales and/or
profitability. Interaction between the marketing department and the market researcher
(or organization) should come up with a clear definition of the problem or situation. It
is important to stay in this stage until both the organization and the researchers are
quite clear and agreed upon what they want from the research, and how it will be
used.

Definition of research objectives

The next step in any research is to clearly define the objectives of the study. If the
objectives are stated too vaguely or wrongly defined, then the research results may be
useless or even misleading to the organization.
Care must be taken in determining exactly what information is needed to assist
decision making and setting objectives which ensure that the need is met.

Research design
The research has to be designed in order to meet the objectives of the study. This
involves decisions about the types of data, the data collection method to be used and
the specific people to be interviewed.
It is important to have a clear statement of the research objectives. The “research
objective” is a statement of the outcomes that we are looking for from the research
project. It may also be that researchers will develop a hypothesis based on past
research and the expected outcomes of the planned research.

The hypothesis is a statement – basically an informed guess – about a particular

problem or set of circumstances. It is based on what we know about the problem
based on past research and any other relevant sources. As the research project
develops and information begins to be produced, then it is possible to test the
hypothesis based upon this information.

There are three types of research to be considered:

a) Exploratory research
b) Descriptive research
c) Causal research.
a) Exploratory research
Exploratory research looks at the general nature of a problem and the factors that
relate to it. The design of this type of research is very flexible. It may be that the
research will be expected to review information from internal company sources or
publicly available information, such as a study of information available on the internet
regarding industry trends or demographics.
b) Descriptive research
This type of research will concentrate on providing an accurate description of the
variables within a situation. An example of this would be where an organization is
looking to target a particular group within the population, say a bank wanting to target
young adults with a new account and carrying out research to determine how often
this group would use ATMs, telephone banking and internet banking. Descriptive
research will demand a high level of prior knowledge and will assume that the
problem is clearly defined. The main task is then to determine the best methods of
collecting and measuring the data.
c) Causal research
This type of research will need a more complex approach than descriptive research. It
is assumed that a particular variable A will cause a variable B. The research must be
planned to either prove or disprove that A causes B.

When designing research, we must ensure that the research techniques are both
reliable and valid. A technique will be deemed to be reliable if it produces very
similar results in a series of repeated trials. However, to be valid, the technique must
measure what it is supposed to measure.

There are two types of data:

1. Secondary data
This is data that is currently available and which was originally collected for purposes
other than the specific research needs at hand. This will include the use of in-company
data, published and unpublished sources such as official statistics, newspapers,
technical journals, past business dissertations and other information available from
libraries. These sources provide a relatively inexpensive supplement to any primary
data gathered.
It is quite common for market research projects to commence with the collection of
secondary data, with the organization's own sources of information often being a good
starting point; for example, sales reports, previous research activity outcomes, etc.
The sources can vary depending on the nature and size of the organization; however,
here are some examples:
• accounting and financial information, including data on sales, expenses and profits,
and may be collated for particular types of customer, geographical area, product type,
etc
• customer feedback – most organizations will have a system to capture feedback
from customers, both positive and negative; in addition, in the financial services
sector, there are the complaints standards laid down by the FSA that must be adhered
to
• comparative information on competitors – again, it is common for organizations to
collate information on the products and services offered by the competition, along
with pricing information.
Again, there are a number of potential sources here, such as government statistics,
the internet, trade associations (in financial services, journals and periodicals, etc.

2. Primary data
This is new data collected specifically for purposes of the research needs at hand and
can take four forms:
a) Observational research
b) Interviewing
c) Surveys
d) Experimentation

a) Observational research
One way to collect primary data is to carry out personal observations in various
situations, such as observing customers’ actions in a bank branch. Video cameras are
often utilised to collate the information. This can help in the design of new bank
branches. When carrying out observation research, direct contact with the respondents
is avoided. Rather the researchers will record the respondents’ behaviour, taking
account of physical conditions and events, therefore the researchers are taking note of
respondents’ actions. This type of research may involve the customers being observed
using the product, or wider services from the organization, such as the premises
mentioned earlier in this section. This type of research is not just limited to the
consumer.
The most common example here is the use of mystery shoppers, who sample the
service of organizations and score the service level received. Data that is collected
through observation may be flawed if the respondent is aware of the observation
process. If it is thought that the presence of a human observer is likely to skew the
behaviour of the consumer, this may be overcome by the use of mechanical
observation devices, such as security cameras in branches to monitor queue
movements and trends. Loyalty cards used by supermarkets are another example of
observational research, where the organization is able to gather in data on the
consumption patterns of customers.
b) Interviewing
This is perhaps the main activity most people associate with marketing research. It
often involves the use of a questionnaire, which is administered by an interviewer or
by the respondent him/herself. It can be completed on the respondent’s doorstep, in
the street, by telephone or through the mail. The format of the questions will vary
depending on the objectives of the research, with very open and discussive questions
being used to probe deeply into customer motivation and multiple choice/yes-no type
questions being used where basic data on customer behaviour is required.

c) Group discussions and extended interviews

These are used in what is called qualitative research and are concerned with gathering
a significant depth of information from a small number of respondents rather than a
significant volume of information from a large number of respondents. Extended
interviews involve obtaining indepth information from one individual at a time. As
their names suggest, these types of interview will tend to be longer and more
discussive in nature. Group interviewing typically involves having a trained
interviewer meet for a lengthy period with eight to ten persons in an informal setting.
The interviewer encourages free and easy discussion among participants, hoping that
the group dynamic will bring out real feelings and thoughts.

Surveys
Similar to interview scenarios with questionnaires usually employed to collate the
information. Whilst some are completed by interview, many are sent by post for self
completion with an increasing number via the internet.

Sampling
When looking at both interviewing and surveys, it is useful to give some thought to
the concept of sampling. As we have already discussed, all organizations have to
operate with limited resources. As a result, it is not possible for them to consult with
every customer or potential customer when they are carrying out market research
activity. The “population” is the number of households or organizations that are of
interest to the researchers. Out of this population, the researchers will select a
“sample” which they hope will be representative of the views, needs, wants, etc of the
larger population. You will be familiar with this concept through your knowledge of
political opinion polls, where a smaller group are surveyed in the belief that their
views are representative of the larger population. Therefore the objective of sampling
is to select representative units from a total population. Sampling procedures allow
marketers to predict buyer behaviour accurately based on the responses from the
sample group.

There are two types of sampling – probability sampling and random sampling.
Probability sampling is where every element within the population has a known
chance of being selected for study, whereas with random sampling all of the units in
the population have an equal chance of being selected to appear in the study. A further
type of probability sampling is stratified sampling where a population is divided into
groups according to a common characteristic or attribute, and probability sampling is
then conducted within each group. This technique may help to reduce some of the
error that may occur as a result of using a simple random sample.
Area sampling involves selecting:
• a probable sample of particular geographical areas, for example particular
streets, and
• households, individuals, or other units within the selected geographical areas
for the sample.
When choosing how to select the units, the researchers may decide to select every
nth house, or they may use a random sampling technique.
The final sampling technique is quota sampling where the population is divided into
groups and participants are chosen at random from there.
It is normal to have some form of controls – normally two or three variables, such as
gender, age, height, occupation, etc. These controls will attempt to ensure that the
sample is representative.

Experimentation
Not a word or a methodology that features too regularly in financial services, it is
more akin to manufacturing industry. The closest it gets is possibly pilot schemes
when new products or delivery mechanisms are tested in a limited manner.

Data collection
The data collection or fieldwork phase of the research follows after the research
design has been finalized. There may be some testing of the research method and
sample before the full primary research phase is undertaken. Interviewers then have to
be trained on the particular subject area and in the skills of encouraging accurate and
thoughtful answers from respondents.

Data analysis and report presentation

Once the data has been collected, it must be interpreted if any meaning is to be made
of it. This interpretation will be easier if there have been clear objectives set at the
start of the process. Normally the first step in analyzing the information is to present it
in tabular format.
After this stage, statistical interpretation will commence. This interpretation will focus
on what is typical, and what information deviates from the average which will give an
indication of how widely responses vary and how distributed these responses are.
However, within interpretation, there is likely to be a degree of judgement used by the
researcher. Once the data has been analyzed, this may cause the original hypothesis to
be accepted or rejected.
The data coming out of the research will then be interpreted. If the results of the
study are deemed to be valid, then the organization will need to take decisions based
upon it. However, the analysis may have shown that one or more of the questions in
the research were flawed and the results should be consequently discarded. Finally
the research results must be reported – usually in a formal, written document. It is
common to find that in these reports, the summary and recommendations of the
researchers are presented at the start of the document because many of the users of
this type of report do not have the time or the inclination to plough through the whole
document.