Market Response Modeling

1
Market Response Modeling

• Response Modeling Basics

2
Response Models

• Aggregate response models
• Individual response models
• Shared-experience models
• Qualitative response models

3
The Concept of a Response Model
Idea
Marketing Outputs

• Sales
• Share
• Profit
• Awareness, etc.

4
Input-Output Model
Marketing Actions Inputs
Observed Market Outputs
Competitive Actions
(2)
Market Response Model
(1)
(4)
(3)
Environmental Conditions
Control Adaption (6)
Evaluation (5)
Objectives
5
Response Function
Max
Sales Response
Response Function
Current Sales
Min
Current Effort
Effort Level
6
A Simple Model
Y (Sales Level)
b (slope of the salesline)
a (sales level
when advertising 0)
X (Advertising)
7
Phenomena
P1 Through Origin
P2 Linear
Y
Y
X
X
P3 Decreasing Returns (concave)
P4 Saturation
Y
Y
X
X
8
Phenomena
P5 Increasing Returns (convex)
P6 S-shape
Y
Y
X
X
P8 Super-saturation
P7 Threshold
Y
Y
X
X
9
Aggregate Response ModelsLinear Model

• Y a bX
• Linear/through origin
• Saturation and threshold (in ranges)

10
Aggregate Response ModelsFractional Root Model

• Y a bXc
• c can be interpreted as elasticity when a 0.
• Linear, increasing or decreasing returns (depends
  on c).

11
Aggregate Response ModelsExponential Model

• Y aebx x gt 0
• Increasing or decreasing returns (depends on b).

12
Aggregate Response ModelsModified Exponential
Model

• Y a (1 ebx) c
• Decreasing returns and saturation.
• Widely used in marketing.

13
Aggregate Response ModelsAdbudg Function

• Y b (ab)
• S-shaped and concave saturation effect.
• Widely used.
• Amenable to judgmental calibration.

14
Aggregate Response ModelsMultiple Instruments

• Additive model for handling multiple marketing
  instruments
• Y af (X1) bg (X2)
• Easy to estimate using linear regression.

15
Aggregate Response ModelsMultiple Instruments
contd

• Multiplicative model for handling multiple
  marketing instruments
• Y aXb Xc
• b and c are elasticities.
• Widely used in marketing.
• Can be estimated by linear regression.

1 2
16
Dynamic Effects
1. Marketing Efforteg, sales promotion
17
Dynamic Effects
2. Conventional delayed response and customer
holdout effects
Sales Response
Time
18
Dynamic Effects
3. Hysteresis effect
Sales Response
Time
19
Dynamic Effects
4. New trierwear out effect
Sales Response
Time
20
Dynamic Effects
5. Stocking effect
Sales Response
Time
21
Aggregate Response ModelsDynamics

• Dynamic response model
• Yt a0 a1 Xt l Yt1
• Easy to estimate.

carry-overeffect
currenteffect
22
Aggregate Response ModelsMarket Share

• Market share (attraction) models
• Ai
• Mi
• A1 A2 . . . An
• Ai attractiveness of brand i.
• Satisfies sum (market shares sum to 1.0) and
  range constraints (brand share is between 0.0 and
  1.0)
• Has proportional draw property.

23
Individual-Level Response ModelsRequirements

• Satisfies sum and range constraints.
• Is consistent with the random utility model.
• Has the proportional draw property.
• Widely used in marketing.

24
Individual-Level Response ModelsMNL

• Multinomial logit model to represent probability
  of choice. The individuals probability of
  choosing brand 1 is
• eA1
• Pi1
• å eAj
• j
• where Aj å wk bijk
• k

25
Logit Model Implications . . .
High
Marginal Impact of a Marketing Action
Low
0.0
0.5
1.0
Probability of Choosing the Alternative
26
Attribute Ratings per Store
27
Shares per Store

• (a) (b) (c) (d) (e)
• Share Share estimate estimate without
  with Draw Store Ai wk bjk eiA new store new
  store (c)(d)
• 1 4.70 109.9 0.512 0.407 0.105
• 2 3.30 27.1 0.126 0.100 0.026
• 3 4.35 77.5 0.362 0.287 0.075
• 4 4.02 55.7 0.206

28
Objectives

• Profit( Sales Margin Costs)
• Sales
• Market share
• Time horizon
• Uncertainty
• Multiple goals
• Multiple points of view
• Others ??

29
Shared Experience Models

• Base the response model on behavior observed at
  other leading firms
• Advisor model
• PIMS model