Summary Marketing Analytics
Lecture 1
Marketing Analytics: A technology-enabled and model-supported approach to harness customer and
market data to enhance marketing decision making.
Main tool → Market response models:
Linking input variables to output measures
What is the impact of marketing decisions on output variables such as sales, market share,
consumer choice etc.
Quantifying marketing problems
Benefits of using market response models:
Better decisions
Improved understanding of how the environment works
Improve the way decision makers work with existing information
Improve data collection
Pinpoint changes in the environment
Levels of decision making:
Market-level data: Strategic data → Where should we put our money?
▪ Compare the relative effectiveness of different marketing instruments e.g. price
promotions, advertising, online advertising forms.
▪ How to determine the allocation of budgets across marketing instruments?
Individual level data: Tactical data → How should we spent our money?
▪ Evaluate how marketing actions influence individual consumers: Which consumers to
approach; how to use consumer behavioral response to improve deployment of
marketing actions.
▪ How to allocate our budget across consumers?
Model building with sequential steps:
1. Purpose: What do you want to measure?
2. Specification: How to measure it?
3. Estimation: Specify the data collection and analysis
4. Validation: How well is the model measuring?
5. Use: Use the model
, 1. Purpose: Which kind of model to use, what is the goal of the model?
Descriptive: simple and summarized statistics (MRM)
Diagnostic: explain a phenomenon and/or relationship between input and output
Predictive: predicts effect a certain input variable will have on the output
Prescriptive: this model will not only predict effects but will also try to find the optimal condition of
the model.
2. Specification:
Inputs: marketing decision variables/independent variables → variables marketers can
control
Outputs: what is the target; dependent variables
Environmental factors: which other factors influence the output; control variables
Specification of the functional form: How do you expect inputs to relate to the outputs?
Express in terms of equation:
➢ Think about subscript → what is your level of observation: t for time; i for individual
➢ Error term
➢ Think about functional form
3. Estimation: availability of data is the factor that determines the model
Specify data needed
Identify statistical technique that can be used
Determine parameter estimates: the change in the response associated with a one-
unit change of the predictor, all other predictors being held constant.
, 4. Validation: how good is the model?
Testing: test the predictive power of the model on a hold-out sample
Validation: do the obtained results make sense?
5. Use:
Use of the model
Continued testing
Updating of the model
Koschmann & Bowman (2018). Evaluating marketplace synergies of ingredient brand alliances
Research Question → Diagnostic: How do (in)congruent functional and emotional aspects of brands
influence market share and revenue for brands in IBA?
Ingredient Brand Alliances (IBAs): They incorporate the physical aspects of one brand into another
brand’s product. It is a type of brand alliance, or co-brand, that features two brands in one product
offering. Combining two brands into one product.
Why form IBAs?
Managers seek to incorporate the meaning and value of one brand (the secondary brand) to increase
the value of the existing brand (the primary brand).
Consumers form two global higher-order judgements of brands:
1. Perceived functional associations: How well the brand functions
2. Perceived emotional associations: How consumers emotionally connect with the brand.
➔ The identification of these two associations arises at the earliest stage of the buying decision
process as consumers think about which brands will provide usefulness through functional and
emotional benefits.
Combining attributes and ultimately associations: Given all the possible attributes of a brand,
consumers integrate information into overall judgements of performance and image. These two
higher-level associations are similar in concept across researchers, denoting two types of broad
benefits, described variously as: tangible vs. intangible; functional vs. experiential; functional vs.
symbolic; function-oriented vs. prestige-oriented; and utilitarian vs. hedonic. These terms reflect
global judgements that a customer has toward a brand regarding overall functional performance and
overall emotional connection with the customer. In bringing global consumer judgments of functional
and emotional associations to the IBA, this study focuses on how these associations combine.
Concept combination theory: It addresses how humans process concepts or objects that are
combined, namely through transferring meaning from one concept to another. Concepts are
presumed to have:
Slots: which store attributes of the concept
Fillers: the attributes that fill the slots
Example: An apple (as a concept) has a color slot that is typically filled with red or green. Additionally,
the concept of sauce has its own slots (texture, color, taste) and fillers. Apple sauce combines the two
concepts of apple and sauce, where the taste slot should borrow more from what apples taste like,
while the texture is more likely traditional sauce than a physical apple.
Congruent associations in IBAs:
H1a: The secondary brand’s functional association negatively moderates the primary brand’s
functional association in IBA performance (supported).
, H1b: The secondary brand’s emotional association negatively moderates the primary brand’s
emotional association in IBA performance (supported).
➢ Congruent associations: Host and partner brands have a common set of alignable attributes
that are combined in an IBA.
➢ Spill-over effect: In an IBA, the secondary brand affects not just the values of the primary
brand, but also gives the primary brand attributes it did not previously have. While brands
are being mindful of whether the associations of one brand spill over to another, IBA finds
overwriting of the primary brand by the secondary brand unlikely. The host brand’s existing
attributes are unlikely to change after including the secondary brand as an ingredient,
because consumer perceptions of the secondary brand should apply to the IBA, but not to
the primary brand.
➢ Degree of overlap: Concept combination theory suggests concepts can be assimilated or
contrasted depending on the degree of overlap. An overlap offers limited benefits to
customers. This is because consumers perceive a diminishing utility of congruent
associations. Even if both brands have positive utility, consumers can still devalue the
combination. Similar features compete for similar roles, increasing a perceived trade-off and
combining brands from a similar superordinate level requires consumers to think more about
why these two brands have combined, inhibiting adoption.
Incongruent associations within IBAs:
H2a: The secondary brand’s emotional association positively moderates the primary brand’s
functional association on IBA performance (not supported).
H2b: The secondary brand’s functional association positively moderates the primary brand’s
emotional association on IBA performance (not supported).
➢ The contrast effect: adding an incongruent hedonic attribute of one brand to the functional
attribute of another brand. It makes a product stand out. Consumers ignore similar features,
focusing instead on the differences.
➢ Positive impressions of incongruencies can occur under two conditions: structural alignment
and goal-derived categorization.
Structural alignment: a process in which individuals make non-combinable concepts
more coherent by actively imagining a way to bridge the two incongruent items. This
means customers will think more actively about an underlying link between the
incongruent associations. But this is not enough, incongruency requires a goal.
Goal-derived categorization: Delivering more value and signaling quality for the
customer drives the combination of structural alignment and goal. IBAs naturally
encourage consumers to draw links between the combined brands, with a goal of
delivering value.
Additional notes: the author assumes linear relationships and that functional and emotional
characteristics have an interdependence (interaction).
Conceptual framework:
1. Functional and emotional characteristics of primary brand influence brand alliance sales
performance.
2. This effect is strengthened or weakened by the functional and emotional associations of the
secondary brand → moderation.
Lecture 1
Marketing Analytics: A technology-enabled and model-supported approach to harness customer and
market data to enhance marketing decision making.
Main tool → Market response models:
Linking input variables to output measures
What is the impact of marketing decisions on output variables such as sales, market share,
consumer choice etc.
Quantifying marketing problems
Benefits of using market response models:
Better decisions
Improved understanding of how the environment works
Improve the way decision makers work with existing information
Improve data collection
Pinpoint changes in the environment
Levels of decision making:
Market-level data: Strategic data → Where should we put our money?
▪ Compare the relative effectiveness of different marketing instruments e.g. price
promotions, advertising, online advertising forms.
▪ How to determine the allocation of budgets across marketing instruments?
Individual level data: Tactical data → How should we spent our money?
▪ Evaluate how marketing actions influence individual consumers: Which consumers to
approach; how to use consumer behavioral response to improve deployment of
marketing actions.
▪ How to allocate our budget across consumers?
Model building with sequential steps:
1. Purpose: What do you want to measure?
2. Specification: How to measure it?
3. Estimation: Specify the data collection and analysis
4. Validation: How well is the model measuring?
5. Use: Use the model
, 1. Purpose: Which kind of model to use, what is the goal of the model?
Descriptive: simple and summarized statistics (MRM)
Diagnostic: explain a phenomenon and/or relationship between input and output
Predictive: predicts effect a certain input variable will have on the output
Prescriptive: this model will not only predict effects but will also try to find the optimal condition of
the model.
2. Specification:
Inputs: marketing decision variables/independent variables → variables marketers can
control
Outputs: what is the target; dependent variables
Environmental factors: which other factors influence the output; control variables
Specification of the functional form: How do you expect inputs to relate to the outputs?
Express in terms of equation:
➢ Think about subscript → what is your level of observation: t for time; i for individual
➢ Error term
➢ Think about functional form
3. Estimation: availability of data is the factor that determines the model
Specify data needed
Identify statistical technique that can be used
Determine parameter estimates: the change in the response associated with a one-
unit change of the predictor, all other predictors being held constant.
, 4. Validation: how good is the model?
Testing: test the predictive power of the model on a hold-out sample
Validation: do the obtained results make sense?
5. Use:
Use of the model
Continued testing
Updating of the model
Koschmann & Bowman (2018). Evaluating marketplace synergies of ingredient brand alliances
Research Question → Diagnostic: How do (in)congruent functional and emotional aspects of brands
influence market share and revenue for brands in IBA?
Ingredient Brand Alliances (IBAs): They incorporate the physical aspects of one brand into another
brand’s product. It is a type of brand alliance, or co-brand, that features two brands in one product
offering. Combining two brands into one product.
Why form IBAs?
Managers seek to incorporate the meaning and value of one brand (the secondary brand) to increase
the value of the existing brand (the primary brand).
Consumers form two global higher-order judgements of brands:
1. Perceived functional associations: How well the brand functions
2. Perceived emotional associations: How consumers emotionally connect with the brand.
➔ The identification of these two associations arises at the earliest stage of the buying decision
process as consumers think about which brands will provide usefulness through functional and
emotional benefits.
Combining attributes and ultimately associations: Given all the possible attributes of a brand,
consumers integrate information into overall judgements of performance and image. These two
higher-level associations are similar in concept across researchers, denoting two types of broad
benefits, described variously as: tangible vs. intangible; functional vs. experiential; functional vs.
symbolic; function-oriented vs. prestige-oriented; and utilitarian vs. hedonic. These terms reflect
global judgements that a customer has toward a brand regarding overall functional performance and
overall emotional connection with the customer. In bringing global consumer judgments of functional
and emotional associations to the IBA, this study focuses on how these associations combine.
Concept combination theory: It addresses how humans process concepts or objects that are
combined, namely through transferring meaning from one concept to another. Concepts are
presumed to have:
Slots: which store attributes of the concept
Fillers: the attributes that fill the slots
Example: An apple (as a concept) has a color slot that is typically filled with red or green. Additionally,
the concept of sauce has its own slots (texture, color, taste) and fillers. Apple sauce combines the two
concepts of apple and sauce, where the taste slot should borrow more from what apples taste like,
while the texture is more likely traditional sauce than a physical apple.
Congruent associations in IBAs:
H1a: The secondary brand’s functional association negatively moderates the primary brand’s
functional association in IBA performance (supported).
, H1b: The secondary brand’s emotional association negatively moderates the primary brand’s
emotional association in IBA performance (supported).
➢ Congruent associations: Host and partner brands have a common set of alignable attributes
that are combined in an IBA.
➢ Spill-over effect: In an IBA, the secondary brand affects not just the values of the primary
brand, but also gives the primary brand attributes it did not previously have. While brands
are being mindful of whether the associations of one brand spill over to another, IBA finds
overwriting of the primary brand by the secondary brand unlikely. The host brand’s existing
attributes are unlikely to change after including the secondary brand as an ingredient,
because consumer perceptions of the secondary brand should apply to the IBA, but not to
the primary brand.
➢ Degree of overlap: Concept combination theory suggests concepts can be assimilated or
contrasted depending on the degree of overlap. An overlap offers limited benefits to
customers. This is because consumers perceive a diminishing utility of congruent
associations. Even if both brands have positive utility, consumers can still devalue the
combination. Similar features compete for similar roles, increasing a perceived trade-off and
combining brands from a similar superordinate level requires consumers to think more about
why these two brands have combined, inhibiting adoption.
Incongruent associations within IBAs:
H2a: The secondary brand’s emotional association positively moderates the primary brand’s
functional association on IBA performance (not supported).
H2b: The secondary brand’s functional association positively moderates the primary brand’s
emotional association on IBA performance (not supported).
➢ The contrast effect: adding an incongruent hedonic attribute of one brand to the functional
attribute of another brand. It makes a product stand out. Consumers ignore similar features,
focusing instead on the differences.
➢ Positive impressions of incongruencies can occur under two conditions: structural alignment
and goal-derived categorization.
Structural alignment: a process in which individuals make non-combinable concepts
more coherent by actively imagining a way to bridge the two incongruent items. This
means customers will think more actively about an underlying link between the
incongruent associations. But this is not enough, incongruency requires a goal.
Goal-derived categorization: Delivering more value and signaling quality for the
customer drives the combination of structural alignment and goal. IBAs naturally
encourage consumers to draw links between the combined brands, with a goal of
delivering value.
Additional notes: the author assumes linear relationships and that functional and emotional
characteristics have an interdependence (interaction).
Conceptual framework:
1. Functional and emotional characteristics of primary brand influence brand alliance sales
performance.
2. This effect is strengthened or weakened by the functional and emotional associations of the
secondary brand → moderation.
