LECTURE NOTES MARKETING MST-21306
Lecture 1: Marketing planning, Marketing models and Brand
choice models
• Why use quantitative marketing models?
o Better business decisions with empirical evidence;
o Marketers are held accountable;
o Businesses generate and collect data.
• What is a model?
o Simplified representation of the world;
o Build to help us understand the world and;
o To make predictions.
Identifying competition:
• Differentiation (What you do to an offering) → Creating tangible or intangible differences
on one or more attributes between a focal offering and its main competitors;
• Positioning (What you try to do to the minds of customers) → A set of strategies a firm
develops to differentiate its offering in the minds of its target customers. Successful
positioning will result in the offering occupying a distinct, important, and sustainable
position in the minds of the target customers.
There are different levels of competition (figure 1). Depending on what level of competition you
identify your most relevant competitors, it will have implications for marketing (figure 2).
Figure 1 Levels of competition
1
,Figure 2 Levels of Competition: Implications for Product Strategy
The Brand-switching matrix identifies our greatest competitors and shows (future) development of
market share. Figure 3 presents the brand-switching matrix for our data:
• Rows → brands owned;
• Columns → brand purchased next time.
Figure 3 Brand-Switching Transition Matrix
Loyalty can be defined by looking at the numbers on the diagonal in figure 3. But these numbers
aren’t a clean indicator of loyalty because they also reflect the differential market shares of the
brands. We can view this switching matrix as a special kind of cross-tab, and apply the chi-square
to it.
Chi-square
Cross-tabulations are familiar as the collection of the frequencies with which the ith level of the row
variable occurs in a sample with the jth level of the column variable. The frequency eij we would
expect, derived from a model of independence, is the row sum times the column sum over the
grand sum: figure 4 (e.g. for the first cell: (54x64)/180 = 19,2).
2
, Figure 4 Expected frequencies
Standardized Residuals
Standardized residuals are standardized values corrected with the market share of each brand.
They are the differences between our data and the expected frequencies: figure 5 (e.g. for the first
cell: (O – E)/ SQR E = (49 – 19.2)/ SQR 19.2 = 6.8):
• Values on the diagonal → brand loyalty;
• Positive values → customers are likely to switch to other brand;
• Negative values → expected value was higher than the observed value which means that
they are unlikely to switch to the other brand.
Standard residuals >1,96 in absolute value are significant (meaningful).
Figure 5 Standardized Residuals
3
Lecture 1: Marketing planning, Marketing models and Brand
choice models
• Why use quantitative marketing models?
o Better business decisions with empirical evidence;
o Marketers are held accountable;
o Businesses generate and collect data.
• What is a model?
o Simplified representation of the world;
o Build to help us understand the world and;
o To make predictions.
Identifying competition:
• Differentiation (What you do to an offering) → Creating tangible or intangible differences
on one or more attributes between a focal offering and its main competitors;
• Positioning (What you try to do to the minds of customers) → A set of strategies a firm
develops to differentiate its offering in the minds of its target customers. Successful
positioning will result in the offering occupying a distinct, important, and sustainable
position in the minds of the target customers.
There are different levels of competition (figure 1). Depending on what level of competition you
identify your most relevant competitors, it will have implications for marketing (figure 2).
Figure 1 Levels of competition
1
,Figure 2 Levels of Competition: Implications for Product Strategy
The Brand-switching matrix identifies our greatest competitors and shows (future) development of
market share. Figure 3 presents the brand-switching matrix for our data:
• Rows → brands owned;
• Columns → brand purchased next time.
Figure 3 Brand-Switching Transition Matrix
Loyalty can be defined by looking at the numbers on the diagonal in figure 3. But these numbers
aren’t a clean indicator of loyalty because they also reflect the differential market shares of the
brands. We can view this switching matrix as a special kind of cross-tab, and apply the chi-square
to it.
Chi-square
Cross-tabulations are familiar as the collection of the frequencies with which the ith level of the row
variable occurs in a sample with the jth level of the column variable. The frequency eij we would
expect, derived from a model of independence, is the row sum times the column sum over the
grand sum: figure 4 (e.g. for the first cell: (54x64)/180 = 19,2).
2
, Figure 4 Expected frequencies
Standardized Residuals
Standardized residuals are standardized values corrected with the market share of each brand.
They are the differences between our data and the expected frequencies: figure 5 (e.g. for the first
cell: (O – E)/ SQR E = (49 – 19.2)/ SQR 19.2 = 6.8):
• Values on the diagonal → brand loyalty;
• Positive values → customers are likely to switch to other brand;
• Negative values → expected value was higher than the observed value which means that
they are unlikely to switch to the other brand.
Standard residuals >1,96 in absolute value are significant (meaningful).
Figure 5 Standardized Residuals
3
