Main Points of each Power Point:
1. What is market research, why is it important
2. How do we define the problem that market research aims to solve
3. What are the Market Research designs, when is it appropriate to use them
4. How can we construct surveys to elicit unbiased, truthful responses
What is market research, and why is it important
- Links consumers to marketers through information
- Information shared within the company, used to develop company strategies
- Helps improve understanding of the marketing process
How do we define the problem that market research aims to solve
- Managerial problem: how can the business address opportunities or
challenges posed by the environment?
- Market Research problem: how do we solve the managerial problem? What
information is required, and how do we get this information
o Problem definition: state the general problem, along with its specific
components
Example: What are factors that influence consumer decisions
(general problem) Brand perception, features desired,
competitor offerings (specific components)
o How to acquire information: talk to managers and decision makers in
the company, or talk to industry experts.
- Differences between managerial problem and marketing research problem:
o Managerial problems:
Focuses on what decision makers need to do
Require action
Focus on fixing the ‘symptoms’ of the problem
o Marker Research problems:
Focus on what information is needed
Require research
Focus on finding the causes of the problem
- Mistakes with problem definition:
o If the problem is defined with too much detail (narrow), it may
prevent the researcher from researching other important components
of the problem
o If the problem is defined with not enough detail (too wide), it may be
difficult for the researcher to decide how to go about finding the
information
What are the Market research Designs and when is it appropriate to use them?
Market research designs are the different types of research
- Exploratory: Used to find information on ‘uncharted territory (ex a
completely new industry or product). Things about the unknown component
are clarified, hypotheses developed
o Example a beverage company looking to make first carbonated
beverage. They explore the industry, competitors etc
- Descriptive: Used to describe (obviously)
o Longitudinal: Observe characteristics of subject(s) many times over a
long period of time
Ex. Looking at the heart condition of a group of people over 30
years
o Cross sectional: Observes the characteristics of subject(s) only once at
a specific time
Ex. Observing the purchasing characteristics of parents the day
before Christmas
- Conclusive research
o Aims to provide definitive information that will cause a decision to be
made
- Causal Research: research that examines 2 variables, whether or not one
impacts the other
o Ex. Does showing a celebrity with our product cause positive brand
associations?
How can we construct surveys to elicit unbiased, truthful responses
- Designing questionnaires, examine the following:
o Which info? How to administrate the survey? Question content. How
you want them to respond (open ended answer or yes and no?)
wording (keep it simple) , questions sequence, physical characteristics
(do they need to do something for the survey?), recruiting message
(tell people why they should take the survey, what its for, give
incentive)
Some factors to consider when designing a survey:
- Types of modes:
o In person, at home, online, through email, mail, phone, or intercepted
in a public area
- When to use a personal method:
o The questions are complex, may need to be clarified
o If there is a need for physical stimuli
o If the data needs to be very detailed
- Situational factors:
o Do you need to have control over the survey (in person = lots of
control, internet = no control
o Are speed and cost factors? If so, online surveys are cheapest and
quickest
- How to measure attitudes
o Attitudes are not visible
o Therefore we rely on self reporting measures – questions asked about
attitudes, beliefs, feelings etc.
- Self reported attitude scales
o Graphic ratings scale: the respondent checks a line that has extremes
of an attitude on each side (ex far left is happy, far right is mad)
Infinite choices (you can mark anywhere on the line)
o Itemized ratings scale: the respondent selects from a limited number
of categories
o Comparative ratings scales: rating multiple things using the same
scale (ex scale is : better than others, same, worse than others, and
you’re comparing restaurants)
o Reverse scaling: reverse the order of the values (ex 1-5? Go 5-1 now)
Used to keep respondents on their toes, forces participants to
pay attention
When analyzing this data, make sure to revert the scale (ex. If
rated 1 on a 1-5 scale, it becomes a 5. 2 becomes 4, etc…)
- ‘Don’t know’ category
o Keeps respondents honest, prevents them from choosing an
untruthful response
The problem with this, is that respondents could get lazy and
just check ‘don’t know’ for everything
Determining the sample size
THINK ABOUT IT: would you make a $1million decision based on a sample of 50
people?
There are a few concerns with sample size:
Form a statistical perspective: is the sample size statistically sufficient? Is it
accurate?
From a managerial perspective: Are the managers confident in the results? Does the
sample size make it convincing?
Basic considerations:
- Desired amount of precision in the info determines the sample size
- Precision depends on the size of the confidence interval
o Ex if your estimates are 50-60 apples per harvest +/- 25, youre not
being very precise about how many apples you will get
Estimating means
- if the population variance (σ2) is known:
- if the population variance is unknown, conduct a pilot study using a small sample,
then calculate s using the data form the study. Assume s = population variance
- ratings scales have different variances, if the population variance is unknown. (see
below)
-
