review class summary Flashcards
Purpose of marketing research
3 functions: descriptive, diagnostic and predictive
Predictive: if then, predict decision outcome
Diagnostic: looking at what the problem is, diagnosing why something is a success, what explains this success, looking at marketing mix
Descriptive: find out more about the problem or what the problem is
Problem definition
Determine the business problem (more broad, describing an issue)
Translate it into a research problem (looking more specific an element of your marketing strategy)
Describe specific research objectives (wont be asked)
New coke case
Need to know this case study
Major takeaways
Pepsi was eating up market share so coke introduced the new coke
Pepsi was aggressively advertising, using celebrity endorsements, comparative advertisement and showed how people preferred pepsi to coke
Coke conducted blind taste test and many other tests and survey, they launched a new product called new coke that replaced the old coke
what was cokes mistakes
they put the coke in glasses instead of cans, they also didnt understand the loyalty of consumers (assumed that consumers only cxared about taste, people are emotionally evolved), they didnt tell consumers that new coke would replace original coke
Second mistake: actually research itslef, they did not adequately simulate the experience of people drinking coke, it matters how you simulate the experience, they gave people the drink in a glass and not a can, the amount also affects reward centres in the brain→ did not simulate real experience of drinking coke (people drink more and in a can)
Third mistake: didn’t tell consumers that new coke was going to replace old coke
* remember the takeaways
Product line extensions
Benefits:
Why would a company launc this: to capture market sections
Risk: cannibalization of your other products if there is not enough differentiation between the products and need good segmentation that you have not targeted , how to solve cannibalization: find new segment
Provide more value to customers when you have a product line extension
New coke as product line extension? Not sufficiently differentiated from old coke so this would not have worked, it offers the same value as the old coke
Coke zero or coke lite are different because they have zero calories and are differentiated from new coke or regular coke s there’s a clear differentiation and targeting a new segment
Coca cola life didnt work because people still had associations with coke being sugary
Have to take into account brand image and associations before suggesting a new value proposition
Brand extension:
goes into a new offering market like if coke made chips, enteering a completely new market
Exploratory
Describes a business problem
Secondary research is used → could give scenario and say secondary research is being used so you know its exploratory
Key to understand it:
May be conducted as part of a problem definition
Flexible and adaptive
Ex: understood root problem j crew
Less to do with customers more to do with other aspects of the business
Uses of exploration research:
Define terms
Clarify problems
Establish research priorities
Descriptive
Main goal: Describing a target market
Does not use secondary research, uses observation and survey
More rigid than exploratory
Exploratory is more describing the problem but not necessarily about the customer could be about rising costs
Want to understand your customer
Causal research
Determine causality
More rigid
Running experiments
Ex: does decreasing sugar content affect sales (sugar is the independent because its being manipulated, sales is the dependent variable and its being measured)
Usually generate hypothesis based on causal research (know this), every time you do causal research, have to have a hypothesis
Option: experiment
Examine differences between control and experimental groups under controlled conditions
making causal claims
When making causal claims, your looking at the relationship between two variables (correlation) and a correlation can be explained by causation but not necessarily, have to run an experiment
correlation= two variables share some kind of relationship, can be positive or negative or zero
Correlation can be caused by one way causality, two way causality, spurious and confound
A confound is when you have a relationship between x and y but caused by variable z that is correlated to both x and y, will have a confound when there is no experiment (not controlled and no randomization) there will be a confound
Superious correlation: no causal connection but just so happens that the variables are related
If specified its an experiment,
there are no confounds or spurious correlation so now is it one way causality (most likely)
If not specified that its an experiment, any 4 can be used, up to us to decide which of the 4 are at play
Experiment
Hypothesis
Independent variable→ being manipulated by experimenter , some element of the marketing strategy like if you change price or shelf placement ex price
Dependent variable → the variable the experimenter measures ex: sales,purchase intentions, willingness to pay, consumer attitudes , whether changing some aspect of the marketing strategy (4 ps) how that affects some variable of interest (usually something to do with sales or attitude)
Need randomization for an experiment
Internal validity:
are the findings due to the independent variable ( did I establish causality) → the way you determine this is through randomization and large sample sizes, sometimes there is n o true randomization which means the study is not internally valid→ hard to establish causality
External validity:
can I generalize the results to another group or another context? Usually field - experiments are externally valid bu lab experiments are not, field experiments at grocery store you are getting a sample very much of people who shop at grocery stores,if you use a lab sampling method you are using a convenience sample usually which is not fully representation, field experiments tend to examine people in the natural settings, larger samples, etc, whereas lab samples tend to be weird samples and convenient samples
Can run a field experiment, use more people, diversification of sample and context are very important
