A simple model to study the Mobile Number Portability (MNP)
impact on dynamic behavi of a two-competitor mobile market:
Stability versus Oscillations
Hamed ShakouriG.
Assistant Professor, Department of Industrial Engineering, University of Tehran
Kargar Shomali St, Faculty of Engineering, Campus #2 , University of Tehran, Tehran, Iran,
P.O. Box 14395-515
Email: hshakouri@ ut.ac.ir
Nasim Gh. Tehrani
PhD candidate, Department of Industrial Engineering, Tarbiat modaress University
NASR Bridge, Jlal AleA hma Srteet,Engineering Campus, Tarit Modaress university,Tehran,
Tran, P.O. Box 14115-111
Email:Nasim.T ehrani@ modares.c.ir
Mohammad Hossein Taheri
Head of Mobile Committee, Teyf Sharif Consultants Company
No. 11, Avval St, Arab-ali St, Khoramshahr St, Sohrevardi-Shomali Ave, Tehran, Iran, P.O.
Box 144497-3941
Email:Taheri@ teyf.ir
Abstract
The mobile market in Iran is experimenting a transition period from monopoly
by the incumbent service provider to a relatively competitive market by
introducing the third operator. This paper describes the result of a research
conceming the dynamisms of helping the chum in the mobile network by the
Number Portability service. The feature adds to the attractiveness of the whole
market and potentially stimulates the players to increase their individual
attractiveness by either increasing their service quality or reducing prices or
both. This dynamism is illustrated by a causal diagram, and a flow diagram has
been constructed. The simple proposed model suffices to explore the
consequences of different scenarios provided by the policy makers, and helps
developing suitable policies to launch this new service. It is shown that
improper settings for the service, including the costs, the time to port and the
flexibility of porting can cause undamped oscillations in the system.
Keywords
Mobile Number Portability, Market attractiveness, Market dynamism,
Regulatory role.
1 Introduction
The results of a survey done by Stefan Buehler et al. [1] have presented that
consumers of telecommunications services in the European countries are more
satisfied with their telephone numbers and require saving them switched
between operators. This need causes uncertainty in deciding on which operator
is the most suitable in providing services, while everyone shows tendency to
specific competitive distinctions in the market.
Recently this uncertainty in selecting a service provider has been changed in
some markets since Number Portability (NP) service has been implemented in
many countries around the world and all of the European Union. By this
service, subscribers can choose a new service provider while retaining their
previous numbers to receive services from the new operators. This facility,
which is now available for the both fixed and mobile phone numbers [2], has
improved the attractiveness of the market, and so has increased the penetration
level of the product in the corresponding societies.
MNP service is implemented in most of the mobile markets of different
countries by regulation and feeling of need of regulatory and finally in
collaboration with active operators of the market. The most important aim of
service providers in provision of MNP is increasing facilitation of customer's
selection and portability of subscribers from one operator to another which
promote competition between market players in retaining their subscribers with
taking advantages of strategies such as improving quality of service. Surveys
show MNP service has important role in facilitating consumers’ selection
process of the best provided services in the market and also promoting
competition between active market operators [1].
In Japan, since the introduction of MNP the intense competitive pressure has
been created in the market and has forced operators to provide their services
with lower prices and to develop stronger and more proactive acquisition and
retention strategies. The impact of MNP on the Japan market was not immediate
but as shown in Figure 1, it intensified the competition resulting in higher
subscriber growth in the market [3].
Finland, where the service was introduced in 2003, experienced a huge rate of
chum within a short period. Figure 2 shows a jump in the chum rate from 10%
to about 50% after NP launch [4]. This case can be counted an extra-ordinary
case regarding NP. The main reasons are summarized below.
Porting the number is very easy. When the customers sing up with a new
service operator the new service operator informs the existing one and the
automated porting takes place within two weeks.
Simultaneously with NP service introduction in 2004, the two major
operators, MVNOs and SPs, heavily discounted the tariffs, so Finnish
consumers were used of the price competition
Network operators have launch their own no-frills service providers
providing basic service on low prices.
The role of the regulator in planning and strongly guiding the operators in
the preparatory work for NP was another important matter. The regulator
led the work until a consensus on the system and business model for the
NP between the operators could be reached. The system is owned and
operated by a separate company, Numpac, which in turn is owned by all
the mobile operators. As the system is state-of-the-art, it can support the
fast number porting mentioned above. The regulator is also keen to
measure the porting time.
Figure 1: Year-on-year subsc riber growth in J apan by quarter, 2002-2007.
1
° “cae
G2 G3 G4 QI G2 G3 Qs OI 2 G G4 QI 2 G3 G4 aI G2 OB as I
02 02 02 03 03 03 03 04 04 04 04 05 05 05 0S 06 06 06 06 O7 OF
Source: TCA
Figure 2: Chum evolution among the largest Finnish mobile operators
50,00 % 7
MNP
40,00 % 1 \
1
I
30,00 % i
i == Elsa
1 TeliaSonera
20,00 % 5.
Maw
10,00 % +
1
0,00 %
Qua2 Q3/02 QO? QI/O3 Q203 Q3/03 Q4O3 QUO4 Qu/Os
Rogerson et al. have collected a valuable benchmark on the countries that
recently offered MNP [5]. Except a few exceptions, most of the markets have
shown more or less similar dynamic and static behaviors: a jump in the chum
rate from an average of about 4-5 % up to more than 10% and then retum to
approximately the initial state after a short period of few months.
The market growth dynamics is deeply studied by J. D. Sterman [6], where he
has discussed about 40 loops in the system, from advertisement and awareness
effects to monopoly power and/or the price war. There are not so many works
studying MNP effect and dynamism. However, particularly Jiayin Qi et al. have
analyzed competition strategies of the two major Chinese mobile telecom
carriers based by system dynamics (SD) modeling. They have implemented a
large model, including the regulatory role, which is used to show how
investment efficiency and product innovation can help the carriers to improve
their competitiveness [7].
In this paper, a simple two-competitor model is implemented to show how MNP
can cause the competitors to change their behavior in order to prevent customer
loss. Moreover, the model is adequate to prove that inappropriate adjustment of
system parameters, either set by the regulator or by the service providers, can
lead to oscillation of the customers between the two (or more) competitors.
2 Different Models of NP Offering
NP could be considered in three different approaches: change of the consumer's
physical location, change of the type of service and change of the service
provider (operator), but replacement and change of operators is the main
approach in offering this service .Thus, with respect to this approach, NP can be
subdivided into three main categories [2]:
1. Fixed NP: the number is moved between two fixed operators.
2. Mobile or wireless NP: the number can be moved between mobile service
providers.
3. Fixed to Mobile NP: the number that belongs to a fixed operator can be
moved to a mobile operator and thus be used as a mobile number and vice
versa.
Figure 3: Categories of number portability techniques.
Number Portability
fixed fixed to mobile mobile
non-geographic geographic local nation-wide
same area area independent
Moreover, each category can be subdivided based on its relatedness to region; if
the subscriber replaces his operator with one that provides services just in the
same region as the previous operator this kind of NP is called Local NP, and if
the number ported to another operator irrespective of subscriber it is called
Nation Wide NP.
3 The State of NP Offering in Different Countries
In 1997, Singapore was the first country which implemented Mobile Number
Portability (MNP) in the world. In 1999 the United Kingdom and Netherlands
implemented this service before other European countries. Following these two
countries other countries utilized this facility, the list of which is narrated from
[1] in Table 1 in company with other Non-European countries. Only a few EU
member states have not implemented this service yet.
Table 1: List of the counties that offer the MNP service [1].
Year Countries
1997 Singapore
1999 UK, Netherlands, Hong Kong
2000 Spain, Switzerland
2001 Sweden, Denmark, Norway, Australia
2002 Belgium, Italy, Germany, Portugal
2003 Finland, Luxembourg, Ireland, France
2004 Greece, Austria, Slovenia, Cyprus, Lithuania, Poland, Hungary, USA, South Korea
2005 Estonia, Latvia (planned) , Malta (31 July 2005)
Not clear Czech Republic, Slovakia, New Zealand, Japan, Mexico
The widespread implementation and rollout of MNP service took about six
years in Europe. Some European countries delayed the implementation of MNP
for various reasons. For example Germany postponed the work because of the
lack of appropriate technical solution. Also, Austria delayed the introduction of
this service several times while smaller Austrian operators like Tele.ring and
Tele2 were agreed on MNP, the larger operators including Mobilkom and T-
Mobile were uncertain about introduction of MNP. MNP offerings in some of
Non-European countries such as Australia were done with more than 50 months
delay.
4 The Effective Factors in MNP Offering in the Market
The effective factors in choosing MNP service by consumers have key role in
the success of the service provision. Some of these impressive factors are:
e Number porting costs
e The required period for porting number
e Competitive environment of the market that is affected by quantitative
specifications such as the tariffs and also by quality of offered services
The surveys were done in the European countries have shown that porting
charges and pricing for this service are the most impressive factors among the
others [1].
4.1 Number Porting Costs
High porting costs is an obstacle in adoption of this solution while low costs
(lower than 20 % of operator's monthly revenue) can be effective in increasing
adoption of that [4].
Given to the special effect of this factor in the churn rate of MNP in the market
European union's universal service directive has obliged the member states to
provide the service with cost oriented prices. Thus most of the member states
have to implement regulations that should prevent porting charges to exceed
costs. In Finland MNP service is provided free due to the high competitive
situation of the market. But in countries such as Ireland and Spain operators
are not allowed to set charges for this service and it is an obstacle in
developing porting numbers service [1].
4.2 Speed of porting (the time-to-port)
If the time to port be too long, consumers will put-off using the service. For
consumers 2 days time-to-port is too long. In fact, for some consumers porting
the number within an hour or two is very important .although there isn't any
evidence from international experience to show that reducing a time-to-port
will improve MNP take-up. Y et, on the one hand by reducing porting time, the
cost of MNP implementation would increase. For instance, the cost of porting
a number in Irish system (that has the shortest porting time) is far more than
other implementations and almost five times more expensive than porting cost
in Hong Kong [4].
Speeds of porting and porting time depend on two factors: technical porting
systems and the willingness of networks to speed up the porting process. As a
matter of fact, none of the donor or receiver network tends to resolve technical
problems of porting systems quickly and try to neglect to find adequate cost
solution for this service [1]. One of the other obstacles to rapid porting is the
notice period that has been mentioned in the contract with the donor operator;
in this period the donor operator must be notified before porting a number.
Hence subscriber should wait to end this period for releasing the number by
the donor operator [1].
Although, many countries have utilized manually operated porting process in
the beginning of MNP implementation, nowadays most countries have been
using automatic porting systems. While the speed of porting has been reduced
considerably, there are still some problems in the way of rapid porting [1].
4.3 Competitive Environment in the Market and Quality of Offered
Services
Given that the main purpose of MNP offering is promoting competition
between market players [1], more undesirable competitive situation of market
would have more effect on increasing subscribers’ take-up of porting number
from an operator with undesirable service quality to one that provides highly
desirable service quality. Notwithstanding that the factor has an evident role in
the system; more explanations will be given on this factor in the following
sections.
5 MNP Implementation in Iran Mobile Market
Most of the surveys on MNP provision have been done after offering the
service in the target market. Examining the effect of MNP provision on the
target market and effective factors in customer take-up of MNP service have
been the main purposes of the surveys. But we have been modeling and
examining the effects of MNP provision while this service has not been
offered in Iran yet. Iranian CRA is in the decision making stage on offering
this service, considering technical and legal requirements of it. In other words
there is not any visible behavior and information about market of the service;
hence to facilitate deciding process of Iranian CRA on this service provision,
we have applied system dynamic modeling by examining the surveys and
considering effective factors in the provision of this service and to show
dynamic results of this service on the market offering. Considering natural
benefits of system dynamic modeling, Iranian authorities could deciding more
appropriate on technical and legal framework of the service through changing
different input factors and examining various outputs.
5.1 Overall dynamism
Exploring the dynamism of market and the customer reaction, the overall
dynamism is illustrated in the figure 3. This figure shows the casual loop
responsible for customer decision making about choosing different operators,
and churn. This decision is made as a result of a cost-benefit analysis. As
shown in the diagram, three factors - cost, quality and the variety of the
provided service - in comparison with other operators form this analysis. As
customers decide based on their perceptions and mentality about the real value
of these factors, delay is unavoidable in these loops, illustrated by double lines
on the links. What makes customer decision making meaningful, is the main
purpose of this model, which is churn.
Figure 4: The causal diagram of the main dynamism.
Flexibility to Portt—————— Cost of Porting
a
i os Cost/BenefiT.
Price Difference weg Analysis i,"
Comparitive Chum
Service Variety
5.2 Flow diagram
We added some assumptions to make it possible to model, and quantify the
described loops in Figure 4. As all 3 loops have the same dynamic, and also
because data gathering about the price is simpler than conceptualization and
quantification of other variables (no index has been defined for the overall
quality of service provided by the operators, and also there is no consensus on
the service variety) we are going to model price factor. If we want to consider
the rest, we should define a weighting method to combine all 3 factors in one
index (for the cost-benefit analysis) we also assume that there are only 2
competitors in the market.
As described in the above loop, the price of one operator (tariffs level) ,or in
the more general term, the overall costs and benefits of using a service is
compared with the second service, and the gap is calculated by the costumer.
These costs also include the porting costs. If the gap is considerably positive,
then customer is willing to change the operator and the overall churn rate
increases. It means that one operator is providing the lower quality service, or
with the higher price, or the less variety, and is losing customers. A nonlinear
relation can be considered that describes this fact which is depicted in Figure
5,
Suppose that a percentage of customers decide to leave operator 1, because of
lower overall service. Decision making and implementing this decision is in
company with a predictable delay, which can differ for customers of the two
operators. According to the benchmarks, the delay varies from 1 week to 2
weeks.
There is also another delay associated with the detection of loosing customer
for the operator, and then for the appropriate decision making and the related
policy. In order to show importance of the other parameters, the reaction of the
operators is assumed to be fast enough. Hence the delays are considered to be
about 1 week. This delay is much more for the intangible factors such as
service quality and service variety, which leads to more stability. We have
used “Change Speed k’ variables, to show this behavior. We have also added
the effect of market price by assuming a maximum tariff to the price setting
formula, which is controlled by the regulatory. When an operator attracts more
subscribers, it can increase its price up to this ceiling tariff to attain more
revenue.
5.3. Simulation and the results
Suppose that in time 0, (the service provision time) the number of customers for
each operator is the same, and both operators have the same quality and variety.
But the tariff level for the second operator is 20% higher than the other (as it is
currently allowed by the regulator in Iran). As a result, the customers of this
operator will tend to leave the operator, considering the overall cost (cost of
porting and tariffs and also the time needed for requesting churn and doing it
technically)
For evaluating the number of customers we could use (1), which is also depicted
in Figure 6.
Customers like to churn to 1 = Max(0, (1-EX P(Churn Speed 1*(1 - Total
Benefit 1)))*Customers 2) (1)
Figure 5: The function for “Customers like to chum to k” [%] versus “Total Benefit k”,
based on the formula (1)
Figure 6: The flow diagram for the two-competitor model
chimto 2
\
Bifect Chum Speed 2
_ < \
Change Speed
‘Tariffs of Porting” > Porting Costs <4 Flexibility to Port
Cait
~ ge Op! Tanifis| /
; sale ot Price ra Chum Speed 1 Delay Time 2
al ri
Total Benefit 1 __~ ime 1
eisai ies Delay time ¥
~ Z > \. chan {
Level
ea x Chum Rate 2<———_| pDeay1
nxpuiit ie
Desired Tariff 1 Desired Tariffs 2 Customers 2 op. —— ae
ae i \ a Chnage2
Maximum tariffs Opl
7 Changel
Customers
By assuming a specificied speed for changing the operator (which is characterized by the
flexibility of the system due to the churn and the availability of service provision centers)
the conclusion is that if the percepted value of churn is up to five times more for one
operators, it would loose all of customers. “Total benefit”, as defined in the model, is the
ratio of benefit (price gap between the two operators) to the churning tariff. This process
would take place in an acceptable time frame (such as a few weeks). The “Churn speed”
and “Time delay” are necessary for modifying the decision making speed and the action
speed.
By changing the customers’ number, moving from the lower benefit operator to the
higher one, another mechanism would be activated for the operators. The second operator
tends to decrease the price to attract the customers (bridges the gap in market share) and
the first operator would move toward the maximum acceptable tariff to increase the
revenue. Based on this behavior, there could be a formula for defining the desired tariff,
based on formula 2.
Desired Tariffs = Min(Maximum tariffs, Maximum tariffs/Customers Ratio) (2)
Because the number of operators in this model in assumed to be 2, “customer ratio” is
simply defined by the ratio of number of customers. Changing tariffs is made upon the
speed (agility of operators) to change their price, and is proportional to the desired tariff,
as shown in formula 3.
Tariff Change = (Desired Tariffs - Op. Tariffs Level)*Change Speed (3)
If the prices for operator 1 were low, now it should be increased, and since the ope
operator would decrease the prices, the situation could be vice-versa in a time period. If
the parameter setting is inappropriate, this structure could cause an unrealistic
fluctuation. However, as in the most markets, after 2-3 oscillations it would lead to some
kind of equilibrium. By changing parameters, a parameter set, which is shown by (4)-
(10), is found for the appropriate behavior of the system leading to a stable condition.
This behavior, as the base case, is shown in Figure 7.
Change Speed1=0.4 (4)
Change Speed2=0.7 (5)
Delay time 1=1 (6)
Delay Time 2=1.5 (7)
Flexibility to Port=0.5 (8)
Maximum tariffs=1.5 (9)
Tariffs of Porting=0.7 (10)
Figure 7: Behavior of the system, in equilibrium due to launching MNP att-0, set by values
shown in formula 4-10 (the base case)
Current Current
Op! Tariffs Level Op? Tariffs Level
2
2
17 1.65
14 13
li 0.95
0.8 0.6
Tariffs change 1 Tariffs change 2
0.4 0.4
0.2 0.2
0 Nj. | ot | 0
-0.2 -0.2
04 -0.4
0 13 25 38 50 0 13 25 38 50
Time (Week) Time (Week)
Current Current
Opl Op2
200 200
170 165
140 130
110 / \ / 95K ‘ia \
80 60
Change1 Chnage2
20 20
10 10
Se ee
-10 -10
-20 -20
0 13 25 38 50 0 13 25 38 50
Time (Week) Time (Week)
In this run, the churn rate between operators is as shown in Figure 8. Note that the
variable is indeed the absolute churn due to the NP. Normally there is a small churn rate
which may be added to this value.
Figure 8: Changes in the chum rates (the base case)
Churn Rate 1
0.1
0
0 5 10 15 20 25 30 35 40 45 50
Chum Rate 1 : Current
Churn Rate 2
0.2
0.1
0 5 Jo 15 20 2 30 35 40 45 50
Time (Week)
Chum Rate 2 : Current
As shown in Figure 7 and Figure 8, at first 10% of customers are chuming from the 2"
operator to the 1“. Then the mobility is reversed and reaches to 8% from the 2™ operator
to the 1“. this rate oscillates until it diminishes .It is obvious that this churn facility
benefits for the 1* operator, because of the lower prices. This operator not only has more
customers, but also could increase the prices.
Another analysis at this stage concerns the change of tariff. If we decrease the porting
tariff by about just 10%, we would see a dynamic oscillation behavior in the system, as
shown in (11).
Tariffs of Porting=0.625 (11)
As shown in the Figure 9, in this situation a competition starts between operators. They
try to attract customers by decreasing the costs, and when it is done, they start to increase
the prices for gaining more income, that makes customers to move to the other one. This
makes an unstable market that is not desirable from the regulatory point of view. Then a
restricting rule could be proposed to improve this situation. This simulation shows how
the market would be sensitive to parameter misalignment, although is somehow
exaggerated in this case. We note that for existence of many other nonlinear saturating
and/or tight factors, the oscillation will not happen so easily and the real frequency of
oscillation cannot be that much fast. However, it is quite important to mention that
similar improper behavior (undamped oscillations) may happen if any of other
parameters, like the speed or the flexibility of porting is set to improper values.
Figure 9: Behavior of the system by applying a misaligned parameter (inappropriate policy)
Current
Comparitive Price
2
1.65
1.3
0.95
0.6
Op! Price
2
1.75
15
1.25 yo NAAN
1
Op2 Price
2
1.75
15
1
0 13 25 38 50
Time (Week)
6 Conclusion and further research
As described in the previous sections, most surveys concerning the provision of MNP are
done after launching service in the market. This model has been built for providing a
perspective for investment and policy design about this service, before launch it in the
market. As we have not found a similar model in the literature, we acclaim that this
dynamic model is the first SD model conceming the customer behavior in this market
due to the MNP facility, and could be a basis for future development. It is shown by the
model how any false policy, which can be represented by a misaligned parameter in the
model, can cause undesired dynamism for the system, no difference either the regulator
has taken that weak policy or the operators.
This model is the output of a research project for the Iranian regulatory agency, which
could act as a decision aiding tool. This model has been validated by simulating many
scenarios provided by the managers, and then showing the predicted behavior. The
parameters have been tested by extreme point tests, and then their behavior has been
observed in the regular scenarios. A fter validation, some certain scenarios for launching
this service in Iranian market has been designed by the authors, and presented to the
agency. The policy makers had the opportunity to see the results of each scenario by
changing the different parameters, and observing the outcomes, so the quality of their
decisions improved (by their opinion).
Although this simple model suffices for the goal explained above, i.e. MNP role in
the mobile market from the regulatory point of view, we mention the following
deficiencies to be improved in the next stages to better model and simulate the market
behavior:
e Currently there is not any real proxy measuring quality of services (QoS) or index
referring to customer experience, but consumers have effective ways of filing
complaints towards the operators' business practices and QoS to both the
Communications Regulatory Authority as well as the Consumer Agencies.
Defining quantitative indices for quality and variety of services, and measuring
these indices for every operator by a customer survey, in a time horizon can help
better quantizing of the model.
e Defining weighting system for combining the variety and quality in the cost-
benefit analysis, which by now just includes the price.
e Developing the model to contain all the active operators in the Iranian market.
e Focus more on the investment part, including the investment barriers and also the
technological abilities of each operator.
Considering the target market of each operator, where there is obvious difference
between majority of each operator’s customers, and considering the different
customer behavior for each sector which would affect the policy making and
revenue generating model of each operator.
We recently noticed that some of these are well done by Jiayin Qi et al. in [7].
7 Acknowledgments
This paper is an outcome of a research project done by Teyf Group for the regulatory of
the mobile communication sector in Iran and is supported by founds from Teyf.
8 References
[1]S. Buehlera, R. Dewenterb, J. Haucapc, “Mobile number portability in Europe”,
Telecommunications Policy, 30, 385-399, 2006.
[2]G.N. Prezerakos, S.E. Polykalas, “Maximizing the adoption of fixed number
portability within the EU: An empirical analysis”, Telecommunications Policy,
31,179-196, 2007.
[3]S. Seong,; “Japan’s mobile price war”, Ovum, September 2007.
[4]T. Smura, “Mobile Number Portability-Case Finland”, Mimeo, Networking
Laboratory, Helsinki University of Technology, 2004.
[5]D. Rogerson, M. Holland, N. Griffiths, “Mobile Number Portability - an
international benchmark”, Ovum, Project CLM42, Version Final, 2005.
[6]J. D. Sterman, Business Dynamics, McGraw Hill, 2000.
[7]Jiayin Qi, Ling Li, Hua Ai, “A system dynamics approach to competitive strategy
in mobile telecommunication industry”, Systems Research and Behavioral Science
Volume 26, Issue 2, p.p. 155 - 168, 2009 John Wiley & Sons, Ltd.
