Transcript Document

Business Application of
Agent-Based Simulation
Complex and Dynamic Interactions of Motion Picture Market
SwarmFest 2004
May 11, 2004
이 승규 Seung-Kyu Rhee
이 원희 Wonhee Lee
1
Movie: The Product and
the Market

Movie



The Product



Is naturally a new product and
Has short life-cycle from one week to several months
With huge initial investment and
High uncertainty of the market performance  Highly risky business
The Market

Constituents of the movie supply chain




Consumers in complex social network



From a writer with an idea
To theater managers with screens to allocate and
Everybody in between
Local and central information
Preference and constraints
Competing movies and substitutes
Rhee and Lee: ABS for Movie Industry
2
Movie: The Decisions

Given a movie to sell

A distributor has to decide




Focus of this
paper
How much marketing budget to spend,
When to release it,
How many screens to secure, etc.
The decisions should be based on



Feed-forward
The projected market performance,
Which, in turn, would be influenced by the decisions
themselves and
Many other uncontrollable factors, notably the early
performance of the movie itself.
Rhee and Lee: ABS for Movie Industry
Feedback
3
The Problem

How is the market going to respond to




Various supplier’s decision alternatives under
Various market conditions with competing movies and
The communication dynamics about the movie quality
?
Adaptive reactions of competitors and myself


What-if analysis is critical, but
It is only possible with detailed knowledge of the dynamic process
Rhee and Lee: ABS for Movie Industry
4
Existing Research



Ranges from simple statistical forecasting models to a
complex dynamic Markov chain model with
behavioral parameter estimation
Some agent-based models have been proposed to
describe the near-chaotic market behavior in terms of
market share change
To our knowledge, no existing model is
comprehensive enough to be useful for decision
makers in motion picture industry
Rhee and Lee: ABS for Movie Industry
5
A Sample of Existing
Models
Research
Objective
Method
Authors
Characteristics
Limits
•Forecasting before the
release by estimating
parameters with audience
survey
•Empirical test applied to
real cases
•Competition
•Dynamics
Decision
support system
and Forecasting
Interactive
Markov Chain
Eliashberg et al.
(2000)
Forecasting
Queuing
model
Sawhney &
Eliashberg
(1996)
•Estimating function and
parameters
•Competition
•Dynamics
•Lack of explanatory
variables
•Marketing variables
•Low reliability of
results
•Too simple decision
rule
•Competition
•Dynamics
Understanding
system behavior
Agent-based
modeling
De Vany & Lee
(2001)
•Reliability of product
quality and market
performance feedback by
Information cascading
perspective
Finding major
variables
Empirical
study
Bagella &
Becchetti (1999),
De Vany & Walls
(1999)
•Finding important
variables
•Comparison of coefficient
between variables
Rhee and Lee: ABS for Movie Industry
6
Issues Covered in
Literature
Competition
Movie
characteristics
Marketing
(advertising,
distribution)
○
○
Eliashberg et al. (2000)
Jedidi et al., (1998)
○
○
Lampel et al., (2000)
○
Zufryden (1996)
○
Mahajan et al. (1984)
○
Linton & Petrovich (1988)
○
Litman & Kohl (1989)
○
○
Sochay (1994)
○
○
Lampel and Shamsie (2000)
○
De Vany and Lee (2001)
○
Rhee and Lee: ABS for Movie Industry
Quality
or
WOM
Market
performance
feedback
○
○
○
○
Prag and Casavant (1994)
Critique
review
○
○
○
○
○
○
○
○
○
○
○
○
7
Challenges to ABM

KISS? Reality?



In agent-based simulation community, there is a tendency to
prefer simple models
From practitioners’ viewpoint, however, it does not help
much to confirm the fact that the market is too complex and
anything is possible (e.g., De Vany and Lee, 2001)
Big question: How real is real enough?

In this paper we expand the scope of the movie market
model by including diverse sources of movie quality
information and competition effect.
Rhee and Lee: ABS for Movie Industry
8
Consumer State
Transition Model
Marketing
Strategy
Movie
Quality
Promotion: advertising effectiveness
Place: distribution effectiveness (number of screens)
Product: theme acceptability and audience consensus to quality
Price: irrelevant
Positive
Spreader
Undecided
Movie
Selection
Neutral
Spreader
Inactive
Negative
Spreader
Preview
Performance
Rhee and Lee:
Box Office
WoM in
Neighborhood Modified based on
Competitive and composite quality evaluation:
Eliashberg et al. (2000)
ABScritique,
for Movie
previewIndustry
audience, and audience consensus to quality
9
Agent in Social Network

ABM v. EBM


Example
Daily update of movie-going probability for each agent
Eliashberg et al. (2000) used aggregated market transition
equations
Number of Neighborhood = 6
Number of contacts = 2
Number of WoM communication = 1
Today’s Agent
Contacts
Incommunicable:
Undecided or Inactive (post-WoM)
Communicable:
positive/neutral/negative spreader
Rhee and Lee: ABS for Movie Industry
10
Rich Microstructure in
Agent Model

Modeling objective

Heuristic approach for better understanding of the market


Initial exploration of diverse variables and parameters




Gross and Strand (2000): Predictive, Explanatory, and Heuristic
Sensitivity analyses under diverse scenarios
Part of bigger model: Production-Distribution-Competition
Toward a commercially useful “Decision Support System”
Choice of rich microstructure

The most salient characteristic of “culture products”


Experience goods: performance seriously affected by social interaction
and human intervention
Model saturation can be determined by diverse experiments and
sensitivity analyses
Rhee and Lee: ABS for Movie Industry
11
The Simulation Process
DIstributor
• Number of movies
• Quality of movies
• Marketing effectiveness
• Frequency of marketing
• Audience of preview
DIstributor
• Preview a movie
Audience
• WoM of preview
audience
• WoM update
• Movie-going decision
• Spreading WoM
Critique
Critique
• Frequency of critique
review
• Consensus to quality
• Releasing a movie
• Ending a movie
(No audience)
Audience
Audience
• Number of neighbors
• Number of contacts
• Frequency of WoM
• Duration of WoM
• Consensus to quality
DIstributor
• Review of critique
Media
• Market performance feedback
(Box office report)
Media
• Frequency of market
performance feedback
Set-up
Pre-release Period
Rhee and Lee: ABS for Movie Industry
Release Period
12
General Parameters
Reference
Range
Baseline Model
Number of movies
De Vany & Lee (2001)
2~10 movies
5 movies
Quality of film
Korean movie industry
High, medium, low quality
High: 1, medium: 3, low: 1
Number of audience
N/A
10,000~20,000
10,000 persons
Number of preview
audience
Korean movie industry
1 – 20 (0.0001~0.002%)
10 persons
Preview period
Korean movie industry
1~21 days
7 days
Marketing impacts
Eliashberg et al. (2000)
0.0 – 1.0
0.5
positive, negative, neutral
Dependent on critique
consistency
0.0 – 1.0
1.0
Critique preference
N/A
Critique consistency
movie-going probability
Korean movie industry
0.01~0.05
0.02
Maximum number of
movie selection
N/A
1~5 movies
1 movie
WoM preference
Mahajan et al. (1984)
positive, negative, neutral
Dependent on WOM
consistency
WoM consistency
De Vany & Lee (2001)
0.0 – 1.0
0.7
WoM: neighborhood
N/A
0 – 10 persons
10 persons
WoM duration
Eliashberg et al. (2000)
0-32 days
21 days
1 – 10 per week
2 per week
Rhee andWoM
Lee:frequency
ABS for MovieEliashberg
Industry
et al. (2000)
13
Signal Parameters
Range
Baseline
Model
Marketing signals
1~7 times in
pre-release period
2 times
Critique signals
1~7 times in
pre-release period
0.2 times
WoM signals
Market feedback
signals
Depend on WoM structure
1~3 times
per week
(depends on
release period)
Rhee and Lee: ABS for Movie Industry
Once a week
Characteristics
Performance
independent centralized
information
Performance dependent
(box office and
showing period)
decentralized
information
Performance dependent
(showing period)
centralized information
14
Model Test Using Real
Data

Test movies




Two week brackets in January, February, and July of 2000
Movies with more than 100,000 viewers
Test with opening market share and final market share
Chi-square test (Chung and Cox, 1994)
N
Q
 (Actual
i 1
i
 Predicted i ) 2
Predictedi
Rhee and Lee: ABS for Movie Industry
?
12 ( N 1)
15
Test Data Set: Actual
Movies in Korean Market
Jan.
(Set 1)
Feb.
(Set 2)
Early
July
(Set 3)
Late
July
(Set 4)
Movie title
Opening
day
Critique
quality
Audience
quality
Marketing
impacts
Opening
box office
Total box
office
Peppermint Candy (Korea)
1. 1
H
H
0.2
6,206
290,276
A Happy Funeral Parlor (Korea)
1. 8
L
M
0.3
6,725
111,837
Fly me to Polaris (Hong Kong)
1. 15
L
M
0.3
10,120
202,840
The Bone Collector (USA)
1. 1
L
M
0.4
13,372
212,564
Stuart Little (USA-Germany)
1. 8
L
M
0.4
16,331
392,933
Happy End (Korea)
Lies (Korea)
The Foul King (Korea)
Samurai Fiction (Japan)
The Beach (USA)
1. 1
1. 11
2. 4
2. 19
2. 3
M
M
M
M
M
L
L
H
M
M
0.4
0.8
0.4
0.4
0.3
13,690
19,035
22,741
14,232
14,231
132,029
307,702
787,412
224,256
187,460
The Messenger: The Story of Joan of Arc (France)
2. 19
M
M
0.4
13,084
220,986
Three Kings (USA)
Dinosaur (USA)
Gone in 60 Seconds (USA)
Bichunmoo (Korea)
2. 12
7. 15
7. 1
7. 1
L
M
L
L
L
M
M
L
0.2
0.8
0.5
0.8
10,060
27,859
21,272
23,835
134,376
554,169
348,710
631,913
Bayside Shakedown (Japan)
7. 22
M
H
0.3
13,496
234,155
The Perfect Storm (USA)
The Patriot (USA)
Nightmare (Korea)
Ring 2 (Japan)
7. 29
7. 22
7. 29
7. 29
M
L
L
M
M
M
M
L
0.9
0.4
0.4
0.2
35,184
18,229
12,801
7,164
508,913
149,415
279,174
106,652
Rhee and Lee: ABS for Movie Industry
16
Test Application to Market
Data: Shapes
30000
25000
반칙왕
비치
쓰리킹즈
잔다르크
사무라이픽션
20000
15000
10000
5000
89
81
73
65
57
41
49
33
25
17
9
1
0
Simulation
쓰리킹즈
비치
사무라이픽션
잔다르크
반칙왕
Actual
Rhee and Lee: ABS for Movie Industry
17
Test Application to
Market Data: Fitness
Rhee and Lee: ABS for Movie Industry
18
Result: Baseline Model
Medium
Medium
Low
Medium
High
High Quality
Medium
Medium
Low
Medium
High
Medium Quality
Low Quality
Rhee and Lee: ABS for Movie Industry
19
Baseline Result: WoM
Depletion

WoM intensity gets weaker along the show duration

Initial audience size and signal accuracy (viewer consensus)
intervene
Medium
Medium
Low
Medium
High
Medium
Medium
Low
Medium
High
Low Quality
High quality
Signal accuracy = 0.7
High Quality
Rhee and Lee: ABS for Movie Industry
20
Analysis: Marketing
Impacts
M
kt
g=
0.
1
M
kt
g=
0.
3
M
kt
g=
0.
5
M
kt
g=
0.
7
M
kt
g=
0.
9
4500
4000
3500
3000
2500
2000
1500
1000
500
0
Marketing impacts positively affect the performance
of good movies, and increase the total market size
Rhee and Lee: ABS for Movie Industry
High Quality
Medium Quality
Medium Quality
Medium Quality
Low Quality
Total Market size
9200
9000
8800
8600
8400
8200
8000
7800
7600
7400
7200
M
kt
g=
0.
1
M
kt
g=
0.
2
M
kt
g=
0.
3
M
kt
g=
0.
4
M
kt
g=
0.
5
M
kt
g=
0.
6
M
kt
g=
0.
7
M
kt
g=
0.
8
M
kt
g=
0.
9
M
kt
g=
1.
0

21
Analysis: Marketing
Impacts
4000
3500
3000
2500
Bad movie’s increased marketing impacts

Bad movies only take the market away from other movies
M
kt
g=
0.
1
M
kt
g=
0.
3
M
kt
g=
0.
5
M
kt
g=
0.
7
M
kt
g=
0.
9
2000
1500
1000
500
0
Rhee and Lee: ABS for Movie Industry
High Quality
Medium Quality
Medium Quality
Medium Quality
Low Quality
8900
8850
8800
8750
8700
8650
8600
8550
Total Market Size
Unstable
M
kt
g=
0.
1
M
kt
g=
0.
2
M
kt
g=
0.
3
M
kt
g=
0.
4
M
kt
g=
0.
5
M
kt
g=
0.
6
M
kt
g=
0.
7
M
kt
g=
0.
8
M
kt
g=
0.
9
M
kt
g=
1.
0

22
Analysis: Marketing
Impacts

Decreasing returns to scale for the marketing impact
increase (inducing initial viewer increase) are
confirmed for both good and bad movies with some
irregularities

But if you have a good movie, then excessive marketing do
not help much due to market information spreads
900
800
700
600
500
400
300
200
100
0
350
300
250
200
150
100
50
0
M
kt
g=
0.
1
M
kt
g=
0.
2
M
kt
g=
0.
3
M
kt
g=
0.
4
M
kt
g=
0.
5
M
kt
g=
0.
6
M
kt
g=
0.
7
M
kt
g=
0.
8
M
kt
g=
0.
9
Increase in the box office of low quality movie
M
kt
g=
0.
1
M
kt
g=
0.
2
M
kt
g=
0.
3
M
kt
g=
0.
4
M
kt
g=
0.
5
M
kt
g=
0.
6
M
kt
g=
0.
7
M
kt
g=
0.
8
M
kt
g=
0.
9
Increase in the box office of high quality movie
Rhee and Lee: ABS for Movie Industry
23
Analysis: Marketing
Signals
If consumers take central marketing information
more seriously (than other quality information), the
market growth potential is seriously impaired
4000
3500
3000
2500
2000
1500
1000
500
0
High quality
M edium quality
M edium quality
M edium quality
Low quality
Rhee and Lee: ABS for Movie Industry
Mk
tg
Mk
tg
Mk
tg
Mk
tg
Mk
tg
si
gn
al
=1
si
gn
al
=3
si
gn
al
=5
si
gn
al
=7
si
gn
al
=9
The number of audience

24
Analysis: WoM Range and
Intensity

5000
4500
4000
3500
3000
2500
2000
1500
1000
500
0
R
W a ng
om
e
Ra =0
n
W
om ge
Ra =1
n
W
om ge
Ra =2
n
W
om ge
Ra =3
n
W
om ge
Ra =4
n
W
om ge
Ra =5
n
W
om ge
Ra =6
n
W
om ge
Ra =7
n
W
om ge
Ra =8
ng
e=
9
Rhee and Lee: ABS for Movie Industry
9200
9100
9000
8900
8800
8700
8600
8500
8400
8300
8200
W
om
ng
e=
8
Ra
ng
e=
6
W
om
Ra
ng
e=
4
W
om
Ra
ng
e=
2
W
om
Ra
ng
e=
0
High Quality
Medium Quality
Medium Quality
Medium Quality
Total Market Size
Low Quality
W
om
Ra
W
om
Increasing WoM signals positively affect the
performance of good movies, and increase the total
market size
25
Analysis: WoM Consensus

4500
4000
3500
Increasing WoM consensus positively affect the
performance of good movies, and increase the total
market size
High Quality
Medium Quality
Medium Quality
Medium Quality
Low Quality
3000
2500
2000
1500
1000
Total market size
9500
500
9000
0
1
2
3
4
5
6
7
8
8500
Total market size
8000
7500
Rhee and Lee: ABS for Movie Industry
Random
7
6
5
4
3
2
1
7000
WoM
consensus=1
8
Random
WoM
consensus=1
26
Analysis: WoM Impacts

By the “action-based WoM” assumption, good WoM
spreads widely, but bad WoM does not
Movie
High quality
movie
Low quality
movie
High
Positive
WoM
Listener
No WoM
Negative
WoM
Listener
Number of audience
3539 (35%)
5834 (58%)
609 (6%)
Number of movie-goer
1,275 (46%)
1,436 (52%)
42 (2%)
Movie-goer ratio
41%
22%
7%
Number of listener
497 (5%)
7393 (74%)
2094 (21%)
Number of movie-goer
71 (7%)
954 (92%)
15 (1%)
Movie-goer ratio
14%
13%
1%
Rhee and Lee: ABS for Movie Industry
27
Analysis: Show Duration
and WoM Accumulation

The accumulated impact of WoM shows the “inverted
U shape,” for the movie-going rate per WoM
(probability) decreases after the peak
Good Movie-WoM & Movie going behavior
1800
1600
1400
1200
1000
800
600
400
200
0
0.45
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
Movie goer
Movie going ratio
Longer show and more WoM 28
0이상 2이상 4이상 6이상 8이상10이상
Rhee and Lee: ABS for Movie Industry
Positive WoM
Listener
Analysis: Competition

Movie mix in the market affects the total market size


Good and bad mix is better than all-average movies
If you have a good movie, then release timing
strategy is critical
Average number of viewers per movie according to competition scenarios
Bad movies
Evenly
distributed
Good Movie
when evenly
distributed
2,668
1,977
2,956
4,924
1,492
1,324
1,084
1,446
2,141
991
913
794
956
1,328
The number of
movies
Good
movies
Ordinary
movies
3 movies
3,027
6 movies
9 movies
Rhee and Lee: ABS for Movie Industry
29
Discussion: Market
Growth

Effects of demand growth



Results from increased population (width) and increased
frequency (depth) scenarios show that diminishing returns
to scale
The width shows bigger effect in simultaneous release
competition
Effects of movie supply and mix

Total market size is positively related to


The number, quality and right mix of movies
Marketing impacts and communication effects interact in
different fashion according to the movie quality and mix
Rhee and Lee: ABS for Movie Industry
30
Discussion: Critique
Debates

Debates




Critique influence (Handel, 1950; Litman, 1983)
Critique influence timing: influencer vs. predictor
(Burzynski and Bayer, 1977; Eliashberg and Shugan, 1997)
Critique and consumer taste correlation and independence
(Wanderer, 1970; Eliashberg and Shugan, 1997)
The model can incorporate the different assumptions
and their consequences


What if critiques are ‘influencer,’ ‘predictor’ or both?
It can be shown that the same results can be obtained by
changing parameters of initial marketing impact and WoM
intensity
Rhee and Lee: ABS for Movie Industry
31
Hypothesis for Release
Competition
(Number of movies)
—
Market Size
+
Competition
Quality distribution
(Number of good movies)
Competitor’s marketing
—
Release
Attractiveness
—
+
Marketing signals
Audience
Characteristics
WoM range/probability/
Duration/consensus
Rhee and Lee: ABS for Movie Industry
+
My Marketing
My Quality
32
Discussion: Competitive
Strategy

Actual competition data
Quality
Jan.
Jan. 2000
2000
LOW
LOW
Marketing MEDIUM
HIGH
Early
Early July.
July. 2000
2000
6
MEDIUM
HIGH
5, 7
3
Quality
MEDIUM
3
Rhee and Lee: ABS for Movie Industry 2
1
MEDIUM
5
4
2, 3
HIGH
11
2
Quality
Late
Late July.
July. 2000
2000
HIGH
LOW
LOW
Marketing MEDIUM
LOW
HIGH
2
LOW
LOW
Quality
Marketing MEDIUM
1, 4
LOW
HIGH
Feb.
Feb. 2000
2000
Marketing MEDIUM
HIGH
MEDIUM
5
HIGH
3
2, 4
11
33
Discussion: Competitive
Strategy

Proposed taxonomy of movie quality and marketing
strategy
Quality
LOW
LOW
MEDIUM
HIGH
Weakest
Weak
Marketing
MEDIUM
Focused
Strong
HIGH
Rhee and Lee: ABS for Movie Industry
Strongest
34
Discussion: Modeling
Issues

The model discussed in this paper is one focusing on the
complex consumer dynamics

The concept of model “Saturation”



When applying agent-based simulation to a real and complex decision
situation, it is more important that every additional variable and agent
should be justified by increased insights and relevance
Heuristic approach
Simplified analysis for central v. local communications



On consumer choice
More empirical evidence is necessary for the model improvement
Acceleration phenomenon (e.g., The Passion of Christ, Taegukgi in
Korea)
Rhee and Lee: ABS for Movie Industry
35
Discussion: Modeling
Issues

Model extension directions for practitioners

Market segmentation and competition



Theater objects


Constraints and theater screen mix strategy
Producer objects



Better consumer choice theory is necessary
Overlapping release strategy
Positive and negative feedback of innovation and imitation
Resource-based theory of accumulating intangible assets
Combining the models for practical applications
Rhee and Lee: ABS for Movie Industry
36
Final Thoughts

ABM as a research method


Naturally lead researchers to think more about the
“dynamics” and “adaptive behaviors” than traditionally
thought to be adequate or acceptable
Implications



We need more theoretical models, and
Empirical data based on new models
Especially in practical application purposes
Rhee and Lee: ABS for Movie Industry
37