HEIG-Résumé
CHAPTER 3 – MARKETING RESEARCH
How do managers develop marketing objectives ? How do they select a target market ?
Managers have to get information. This is the fuel that runs the marketing engine.
Look at figure 3.1 p. 105 – We are now in the second section – Understanding Consumers’
Value Needs. For this, we have to conduct marketing research.
THE MARKETING INFORMATION SYSTEM
The MkIS is a process that first determines what information marketing managers need and
then gathers, sorts, analyses, stores and distributes relevant and timely marketing information
to system users.
The 3 components of the MkIS are:
Four types of data (information) Internal company data, marketing intelligence,
marketing research and acquired databases
Computer hardware and software to analyze data and create reports
Information and the decision makers who use it
Internal Company Data:
The internal company data uses information about the results of sales and marketing activities
(which customers buy which products, in what quantities, at what intervals, how many
products in stock, …).
With the MkIS is made available to managers, they can serve their customers better (price of
products, …).
Marketing managers can get daily or weekly sales data for a brand or product lines Think
about MIS Portal in UBS which generates reports.
Marketing intelligence:
Firms have to know their marketing environment. The marketing intelligence system is a
method by which marketers get information about everyday happening in the marketing
environment. It includes in fact information about the environment where you can find in
everyday sources: newspapers, trade publications, web, TV, …
The marketing managers may use marketing intelligence to predict fluctuations in sales
due to economic conditions, political issues and special events. They have to predict
consumer’s trends.
Scenarios, future possible situations, are made.
Marketing research:
Marketing research refers to the process of collecting, analyzing and interpreting
data about customers, competitors and the business environment to improve marketing
effectiveness.
Marketing research is used when unique information is needed for specific decisions.
Syndicated research is collected by firm on a regular basis and then sold to other
firms. For ex. GfK (who does consumer panels) sold information to Nestlé about
consumer’s expectations ( = attentes)
1
HEG-Résumé
Custom research is research conducted for a single firm to provide answers to
specific questions. E.g. Nestlé who does a survey.
Acquired databases:
These are external databases such as government database about economical trend, … .
MARKETING DECISION SUPPORT SYSTEM
The system generates regular reports for decision makers on what is going on in the internal
and external environment.
The problem is that different managers want different and specific information because MkIS
generates a report. So a marketing decision support system (MDSS) has been created.
Managers can access MkIS data and conduct their own analyses. They base on the company
intranet for instance.
A MDSS includes sophisticated statistical and modeling software tools. For instance, when a
manager wants to know how the client perceives the company’s brand and the competitor’s
brand, he can create a “perceptual map” / a graphic presentation.
Problem: as we explained, MkIS systems include a lot of information. These databases are
extremely large. To take advantage of the massive amount of data, data mining has been
created. This is a process in which analysts sift through date to identify unique patterns of
behavior among different customer groups (= c’est un processus dont les analystes grace aux
informations identifient les points uniques de comportement des clients parmis les different
groups.
DIFFERENCE BETWEEN MKIS – MDSS
MKIS MDSS
HAVE OUR SALES DECLINES SIMPLY
WHAT WERE OUR COMPANY SALES OF EACH REFLECTED CHANGES IN OVERALL
PRODUCT LAST YEAR ? INDUSTRY SALES OR IS IT AN INTERNAL
CAUSE ?
IF WE CHANGE OUR MEDIA SCHEDULE HY
WHAT ARE THE BEST MEDIA FOR REACHING
ADDING OR DELETING CERTAIN MEDIA
A LARGE PROPORTION OF HEAVY USERS OF
BUYING, WILL WE REACH FEWER USERS OF
OUR PRODUCT
OUR PRODUCT ?
LESS PERSONALIZED MORE PERSONALIZED
WHAT CAN YOU DO WITH DATA MINING ?
Customer acquisition: here the company can for instance propose client cards. They
have to indicate their name, age, address, … the company know the consumption of
the client then the company can really use these data They know which type of
customer responds best to a specific offer and they can send the same offers to non-
customers
2
HEG-Résumé
Customer retention and loyalty: the firm can identify big-spending customers and
send special offers!
Customer abandonment (=abandon): strange !!! a firm wants sometimes customers
to take their business elsewhere because it costs too much for the company. For
example, hooligans who are not desired or a small client.
Market basket analysis: promotional strategies bases on the records of which
customers have bought certain products. For example, Manor will send food
promotions for big clients who are buying a lot of food in Manor. Facebook will target
publicity according to our customs.
STEPS IN THE MARKETING RESEARCH PROCESS
The goal of a marketing research is to help managers make informed marketing decisions and
reduce risks of taking actions which could harm the business.
STEP 1 : DEFINE THE RESEARCH PROBLEM
First, it is important to clearly understand what information managers need. They are 3
components:
Specifying the research objectives What questions to answer ?
Identifying the consumer population of interest What are the characteristics of the
consumer groups of interest
Environmental context internal and external business environment might be
influencing the situation ?
But this is not really simple. For ex. a car manufacturer wants to find out why its sales gave
fallen off last year? Is the right message being sent? Is the firm’s advertising failing to
reach the right consumers? …
STEP 2: DETERMINE THE RESEARCH DESIGN
When marketers have isolated specific problems, the second step is to decide on a plan of
attack – the research design- it specifies what information marketers will collect and what
type of study they will do.
Secondary research: first of all, marketers should ask themselves is the information
required already exists. So the information are called secondary data. There are for
secondary research two sources: internal sources: company reports, salesperson
feedback, customer feedback, … and external sources: published research,
statistics, trade organizations, … The goal of the secondary research is to see if we
can find the information before doing researches (interviews, …).
Primary research: primary data is information collected directly from respondents to
specifically address the question.
PRIMARY RESEARCH - EXPLORATORY (QUALITATIVE ) RESEARCH
Marketers use exploratory research to generate topics for the future to come up for new ideas
or get new opportunities. They studies are less costly!
Here researchers may interview consumers about products, services, ads or shops!
The results of this research are non-numeric results like consumer’s attitudes, feelings, … .
3
HEG-Résumé
The focus group plays a big role. A focus group consists of five to nine consumers that share
certain characteristics.
They can also use projective techniques: when marketers are not sure if people are
expressing their true reactions a projective test is done: the participant must respond to some
object by telling a story about it. --> True reactions can be found !
The case study is a comprehensive examination of a particular firm or organization. It is
often used in business-to-business marketing to learn how the company makes his purchases,
….
An ethnography is very special. Researchers visit people’s homes or participate in real life
consumer activities. For ex. a researcher follows a consumer for one month to know his
customs (= habitudes)
PRIMARY RESEARCH – DESCRIPTIVE (QUANTITATIVE ) RESEARCH
Then you have the descriptive (quantitative) research which probes systematically into the
marketing problem and bases its conclusions on a large sample of participants. Here it is
expressed in averages, percentages or other statistics measurements
Here you to choices to employ descriptive techniques:
1. A cross-sectional design Questionnaire, … it is a collection a collection of
responses from a survey. The date can be collected on more than one occasion but
with other types of respondents.
2. A longitudinal design the same responses of the same sample of respondents but
over time. It means a weekly or monthly questionnaire –> every month or week.
PRIMARY RESEARCH – CAUSAL RESEARCH
Descriptive research can only DESCRIBE a marketplace phenomenon but it cannot tell us
why. For example, does placing one product next to another in a shop mean that people will
buy more of each ? Descriptive research can’t answer through simple observation.
Causal research attempts to understand cause-and-effect relationships. Marketers uses causal
research techniques to know if a change in something (for ex. place beer next to drinks) is
responsible for a change in something else (increasing sales).
So how do they do? Researchers must do experiments that test predicted relationships in a
controlled environment Respondents are brought to a laboratory so that the researcher can
control precisely what respondent should see. for ex. test about products who are projected.
STEP3: CHOOSE THE METHOD FOR COLLECTING PRIMARY DATA
Now the question is : Which way to choose to collect data ?
1. Survey methods: interview/direct contact with respondents
2. Questionnaires: evolutive questionnaire (may evolve from what the respondent says
to previous questions), questionnaire with fixed choices (strongly agree, disagree, …),
telephone interviews, mail questionnaires, face-to-face interviews, online
questionnaires, … .
3. Observational methods: recording the consumer’s behaviour without their
knowledge (facebook, client cards, …) you can have unobtrusive measures. For
ex. a researcher might go to the house and perform a “pantry check” (= contrôle du
4
HEG-Résumé
garde-manger) and counts the bottles in the drinks cabinet ! mechanical
observation method that relies on non-human devices (=procedés) to record
behaviour for ex. a box on the TV that checks what you watch on TV.
DATA QUALITY : GARBAGE IN , GARBAGE OUT (= LA QUALITE DES RESULTATS ET
FONCTION DE LA QUALITES DES DONNEES A L ’ENTREE )
How much faith (=foi) should marketing managers place in what they find out from the
research? Because too often what the consumer says is not the truth. Three factors influence
the quality of research results:
1. Validity = validité = the customer is loyal
2. Reliabilty = fiability = you can get wrong because you ask questions that creates
errors by biasing people’s responses or people cannot understand well the questions.
3. Representativeness = resprésentativité = the sampling (= l’échantillon) represents
the larger group.
ONLINE SURVEYS – ADVANTAGES
LESS COSTLY
FASTER
THE CONSUMER CAN DO THE SURVEY WHEN HE WANTS
STEP 4 – DESIGN THE SAMPLE
After having defined the problem, decided on a research design and determined how to collect
the data, the next step is to decide from whom to obtain the needed data. They are two main
types of samples:
1. PROBABILITY SAMPLING : in this sample, each member of the population has some
known chance of being included in the sample.
The most basic type of probability sample: SIMPLE RANDOM SAMPLING (in this
sample, each member of the population has the same chance of being included in the
study). SYSTEMATIC SAMPLING échantillonnage systématique selon une règle telle
que avant que soient effectuées les mesures, il y a certitude pour certains éléments de
la population d’être choisis et pour d’autres d’être négligé 2ème, 4ème, 6ème, … .
STRATIFIED SAMPLING the researcher divides the population into segments that
are related to the study’s topic. For ex. men like actions films, women like films with
flicks. The searcher will divide into male and female segments. Then respondent from
each of the two segments would be selected randomly in proportion to their
percentage of the population.
2. NON-PROBABILITY SAMPLING : when researchers don’t have time, effort and cost
required developing a probability sample, they choose this option. With this, some
members of the population have no chance of being included in the sample.
CONVENIENCE SAMPLE : individuals who just happen to be available when and where
the data are being collected. For ex. in a city, when you walk and a researcher is
asking you to fill in the questionnaire
QUOTA SAMPLING : it includes the same proportion of individuals with certain
characteristics as is found in the population. For ex. you are studying attitudes of
students in your university: you will find bachelor, master and PhD students it is
like a stratified sampling except that this is less structured and the researcher uses his
own judgment to select respondent. He can say: “Oh, I will choose this one and this
one”.
5
HEG-Résumé
STEP 5 – COLLECT THE DATA
Then the interviews can be done ! But marketers have to be care of the fact that interviewers
do well their job to have the right answers.
STEP 6 – ANALYSE AND INTERPRET THE DATA
Then the researcher assembles the result in a cross-taulation the date are examined by
subgroups such as male and female, children or parents, … . Then conclusions !
STEP 7 – PREPARE THE RESEARCH REPORT
Then a report has to be prepared. It has to tell clearly and concisely the readers what they need
to know in a way that they can understand. It should include:
- Executive summary
- Description of the methodology
- Tabulations, cross-tabulations
- Limitations of the study (no study is perfect)
- Conclusions and recommendations
6