Page 28 - 33-2

findings remain valid with dependency between the two events (except that the MCN
findings remain valid with dependency between the two events (except that the MCN
company’s profit will be affected by an exogenous amount).
findings remain valid with dependency between the two events (except that the MCN company’s profit will be affected by an exogenous amount).
6.2 Difference in the Effort Exerting Costs
company’s profit will be affected by an exogenous amount).
6.2 Difference in the Effort Exerting Costs
In our basic model, it is assumed that the effort exertion cost is for both
�
�
6.2 Difference in the Effort Exerting Costs
�
In our basic model, it is assumed that the effort exertion cost is for both
creators, where is the effort level and is a common exogenous parameter. With � �
�
In our basic model, it is assumed that the effort exertion cost is for both creators, where is the effort level and is a common exogenous parameter. With
�
�
this assumption, the two creators are identical in the cost of exerting efforts and
�
creators, where is the effort level and is a common exogenous parameter. With this assumption, the two creators are identical in the cost of exerting efforts and
Optimal Advertorial Allocation and Contract Design of a Multichannel Networks Company on Video Sharing
different only in their attractiveness (modeled with the two different values of and
Platforms �
this assumption, the two creators are identical in the cost of exerting efforts and different only in their attractiveness (modeled with the two different values of and
). It is admittedly true that in some cases the two creators may also be different in
�
�
different only in their attractiveness (modeled with the two different values of and ). It is admittedly true that in some cases the two creators may also be different in
6.2 Difference in the Effort Exerting Costs
their effort exertion costs. To model this, we now assume that the effort exertion cost
�
�
k
In our basic model, it is assumed that the effort exertion cost is for both creators,
2
e
). It is admittedly true that in some cases the two creators may also be different in their effort exertion costs. To model this, we now assume that the effort exertion cost
2
is
for the type-L creator, where ≠ . Note
for the type-H creator and
� � � � � � �
where e is the effort level and k is a common exogenous parameter. With this assumption,
�
�
�
�
their effort exertion costs. To model this, we now assume that the effort exertion cost is � � � � for the type-L creator, where ≠ . Note
the two creators are identical in the cost of exerting efforts and different only in their for the type-H creator and
�
�
that we do not assume > or < as either way is possible.
�
�
�
attractiveness (modeled with the two different values of β and β ). It is admittedly true �
�
�
�
�
L
H
is � � for the type-H creator and � � for the type-L creator, where ≠ . Note
�
�
that we do not assume > or < as either way is possible.
�
�
� that in some cases the two creators may also be different in their effort exertion costs. To � � � �
�

Under this extended setting, the analysis may still be done following the same
2
model this, we now assume that the effort exertion cost is for the type-H creator and

e
k H
that we do not assume > or < as either way is possible. Under this extended setting, the analysis may still be done following the same
2
�
�
�
backward induction procedure we used for our basic model (except that the technical
�
2
e
k L
for the type-L creator, where k ≠ k . Note that we do not assume k > k or k < k as
2
H
H
L
H
L
L
either way is possible.
Under this extended setting, the analysis may still be done following the same backward induction procedure we used for our basic model (except that the technical
condition < for the homogeneous cost case is replaced by <
� �
� �
Under this extended setting, the analysis may still be done following the same
� ). With the equilibrium revenue sharing ratios derived, the following
backward induction procedure we used for our basic model (except that the technical condition < for the homogeneous cost case is replaced by <
backward induction procedure we used for our basic model (except that the technical
� �
� �
� �
condition β β A<2k for the homogeneous cost case is replaced by ). ). With the equilibrium revenue sharing ratios derived, the following
� � affect the MCN
proposition helps us understand how the new cost coefficients
condition < for the homogeneous cost case is replaced by < �
H
L
� �
� �
With the equilibrium revenue sharing ratios derived, the following proposition helps us
� ). With the equilibrium revenue sharing ratios derived, the following proposition helps us understand how the new cost coefficients affect the MCN
company’s choice of the revenue sharing ratios, for which we say ( , ) is for
�
� �
understand how the new cost coefficients affect the MCN company’s choice of the revenue
�
�
�
proposition helps us understand how the new cost coefficients affect the MCN company’s choice of the revenue sharing ratios, for which we say ( , ) is for
I
I
sharing ratios, for which we say ϕ (k , k ) is for the type-H creator and ϕ (k , k ) is for
�
H ) is for the type-L one.
�
L
L
L
H
the type-H creator and (
�
�
�
H ,
�
�
�
the type-L one.
company’s choice of the revenue sharing ratios, for which we say ( , ) is for the type-H creator and ( , ) is for the type-L one.
�
�
�
�
Proposition 6. Suppose that < � . Given any values of and such
�
Proposition 6: Suppose that . Given any values of k and k such � � �
H
L�
�
� �
� �
the type-H creator and ( , ) is for the type-L one. H L I H L Proposition 6. Suppose that < � . Given any values of and such
�
I
that k ≠ k , ϕ (k , k ) increases in k and decreases in k , and ϕ (k , k ) increases in k
L
H
H
L
H
�
�
H
�
that ≠ , ( , ) increases in and decreases in , and ( , )
L �
�
�
and decreases in k . � � � � � � � � � � � � � � �
L
Proposition 6. Suppose that < � . Given any values of and such that ≠ , ( , ) increases in and decreases in , and ( , )
�
�
increases in and decreases in
� � .
�
�
�
�
�
�
�
�
�
�
�
� �
�
�
�
According to Proposition 6, the revenue sharing ratio for a creator decreases in the and decreases in .
that ≠ , ( , ) increases in and decreases in , and ( , ) increases in � �
�
�
�
�
�
�

�
�
�
�
�
�
creator’s cost coefficient but increases in that of the other creator. The intuition is the
increases in and decreases in .

following. Let’s say the cost coefficient of the type-H creator, k , has increased. This is
�
�
H
going to drive the type-H creator to exert a lower effort and decrease the probability for
27
her to meet the threshold. As the MCN company realizes that the efficiency of sharing

I
revenue to induce a high effort level is reduced, its optimal response is to cut down ϕ 27
H

(k , k ) to avoid giving out some inefficient share to the type-H creator. On the contrary,
27
L
H
it becomes relatively easier for the MCN company to induce the type-L creator to exert a
high effort. The revenue sharing ratio ϕ (k , k ) should thus be increased to capture the
I
H
L
L
additional efficiency.
I
I
It is interesting to compare Proposition 6 with Proposition 2. While ϕ and ϕ are
H
L
derived in Proposition 2 by assuming k = k = k, the first-order derivatives of ϕ and
I
H
H
L
ϕ with respect to k are quite messy and do not generate insights regarding how the cost
I
L
coefficient affects the revenue sharing ratios. By splitting the cost coefficient into two
20