Bookmark & share

By Liu Nanjie

Telecom transformation is a popular topic at present. Transformation, as such, is by nature structurally subversive, thus, problems that will arise in conjunction with it are part-and-parcel of the process. During the past 130 years, telephone call service has been the mainstay of the entire telecom industry, and it still remains the corner stone of telecommunications. However, this service will no longer be the predominant form of service for telecom operations in the future. Besides voice, there are many other kinds of non-voice services such as: Internet, digital media, entertainment, as well as E-commerce. Hence, a lot of unprecedented issues will naturally arise as a result of this kind of diversity in services.


Changes in Network Profitability

The telecom industry, in-and-of-itself, depends on certain resources for its existence and continued operation. One principal form of resource is the infrastructure, which will take quite a long time, like tens years or more, before a return on investment is expected. The network used for telephone service is composed of two elements: switch and transmission equipment. Voice revenue is directly related to the attributes of switch and transmission equipment (Erlang, BHCA, λ and etc.). Therefore, for these types of operators, no matter what kind of operator it is, ILEC, CLEC or CAP (Competitive Access Provider), the profit model is clear and precise. The majority of investment in the telecom network goes into purchasing network elements like: transmission equipment, switch & access equipment, ports, network management system, gateways, signaling equipment, the operation & maintenance system, billing system, and so on. Hence, the revenue generated from these various elements depends on resources such as: the number defined by E.163/E.164, frequency, bandwidth, λ, port, MoU, and so on. The traditional telecom economy would be in pretty good shape if closer consideration was paid to investment and operation cost & benefit of network infrastructure, which consists of these various network elements. Accordingly, the regulation and accounting strategy, virtual operation and opening-up strategy, as well as the VAS marketing and service strategy, could more easily be defined in the process.

Nevertheless, if the network is mainly composed of packet (IP packet) switch based equipment like: routers, Ethernet switch, MSAN, storage network, media distribution center, BRAS, HSS/GUP, OCAF, a software platform, media server, SIP-AP, service switch, SAN, and IT equipment, then the network's capability of providing service will be greatly enhanced. Nevertheless, as various alternative technologies begin to emerge, packet communication actually causes a loss in revenue for overall voice service. Concurrently, how revenue generated from non-voice service links to the investment structure for new infrastructure is still unknown at present. Some "links" seem to contradict the traditional means of obtaining revenues, thus, making it hard to presently identify the profit point, and even the long-term return on investment in the network.

Looking at it from another perspective, the world has recognized the viability of the Ambient Network being deployed in Europe, as well as the Ubiquity Network in Japan. Thus, further exploration and discussions revolving around these two cases, along with the emerging Internet2.0, will be significant in terms of the economic impact they will have on the telecom industry.


Economic Confusion after Transformation

After the transformation process is completed, what kind of system will new operators be operating on? What will be the primary mode of operation? In other words, how will the telecom operation bring value to customers in the future? In the past, there was only one type of network, in terms of operation. Will the networks of the future be structured similarly, or will they look and operate differently? How will customers be able to differentiate between the network and the service? For instance, if it is the service(s) that will be operated, then there will be issues with how to structure the cost of the service(s). Previously, what people were primarily concerned with was the cost of the network. This is totally different from a cost of service approach.

Will it also be referred to as a telecom operation after the transformation process is finished? If that is the case, then will the industry be monopolized by one or two large companies, which are in turn regulated by the government, much like the agricultural industry, which will account for only a small proportion of the future economy, but it is still perceived as critical national resource.

Most likely the post-transformation telecom service will develop into a comprehensive service, where 90% of the traffic will be non-voice and untraditional type services, and the quality of service will change as well. In other words, operators will be switching from quantitative, to more qualitative types of services. Many kinds of new economic issues will result in the wake of this change as well. For example, what entity or person would be responsible for the input of service capability, and how will service input relate to service revenue? There is also the matter of how billing of services between networks will be settled? These are but a few of the questions that will have to be addressed when the transformation process in completed in the telecom network of the future.

We can divide the network into two layers, based on network function. The first is the pipeline, which can still be perceived in the scope of a monopoly run industry. Although its model of return on investment is quite clear, it will no longer be applicable in the former sense, especially with the emergence of a great deal more of non-voice type services. The second is the service capability layer. With the change in operation subject, the network will also be divided into a core network, bearer network and service network. The Internet will fall into the service network layer (as shown in Fig.1). Here, one very noticeable difference will be that the service value will be out of proportion with the borne traffic, when settling accounts between networks.

For instance, supposing that service settlement (FMC) can be accomplished between Chinavnet (China Unicom) and Monternet (China Mobile). Will it be tabulated by way of borne traffic, or by the value of services? In fact, a "bit" can produce value, whereas a bandwidth of 64K is not even worth a couple of Dollars. For instance, the price for the entire bearer network is relatively low, while the value of services that it bears like, short message is comparatively high. It is also more bit-profitable than, media stream, which is a typical service of 3G and broadband, and therefore, it should have a higher consumption value than short message.

Since the cost of many of these types of services contains a lot of cross-operation cost, one might rightly ask; how will the services be provided, according to cost, or by value? As for the pricing of services, many people may have some understanding about network pricing, but may have problems with the pricing of 3G services, which are orientated to provide broadband and mobility services. Should such services be based on actual costs, or marginal costs? If the net consumption value is the basis for the existence of such services, then how will one be able to amortize the marginal cost? There is also the matter of how best to reflect the ROI of the many CPs, SPs and partners that will be involved? For instance, can the current flat rate fee model (monthly) be used to encourage the creation of services that may have a higher user value?

At present, the value of an Internet enterprise (e.g. Google) is much higher than that of some telecom network companies. Ordinarily, the value of an enterprise should be in proportion to the money put into it, so why is the value of Internet enterprises much higher than that for traditional network companies? For the operators, or the professionals in the telecom industry, issues such as these need to be seriously considered.

When we conduct a 3G economic analysis, the two most frequently considered factors are operation investment and service cost. For the present, if the network is divided into the bearer layer (traditional monopolistic resource-based) and service layer (service network is knowledge-based and capital-based), then the CAPEX and OPEX of the service network will increase when more and more services emerge with stronger profitability. A sudden boost in service revenue, is usually accompanied by another issue: the growth rate of service cost is generally much higher than that of the revenue, but such increase does not come from the bearer network. At this point, a strange IOR curve (as is shown in Fig.1) will occur. So, how are the new OPEX and CAPEX structured and calculated, and how is the overall investment related to the service revenue?

Moreover, there is another important issue that someone may have already considered. That is the issue brought about as a result of the high speed of technical replacement. For instance, the CAPEX of running and operating the traditional network was expected to be viewed as long term asset input. For instance, one could take China Unicom's ISO95 as a case in point: one year after the initial investment, the equipment was replaced by 1X, and then quickly upgraded to EV-DO. It is like the accounting of computers in the early days: even though the depreciation period had been originally estimated to be about five years in the beginning, some hardware, or software in the computer was actually substituted by newer and more advanced products, and so the depreciation period for some computer related components to zero was actually three years, or less. Thus, the actual life cycle for some equipment is in many instances much shorter than the financial depreciation period would allow. Therefore, how will the CAPEX and OPEX change in such circumstances?

In regards to virtual network operation, especially non-voice service operation, a new issue to be faced will be concerned in how best to regulate them. For example, the traditional voice service can be monopolized like voice services reorganized by the U.S. government in the 1950s and 1960s. However, non-voice services such as, media and data services, are capital-based, or knowledge-based services. In order to operate such services, proper regulation (first based on an economic analysis) would become an important issue. Additionally, the concept of virtual network operation should be considered as well, otherwise, it might have a significantly negative economic impact on the entire industry chain, with CAPEX and OPEX already invested in the operation.

It is projected that by 2009, two-thirds of all callers will be mobile subscribers; hence, voice service through a fixed network will decline. Because there are so many technical ways to enable a phone call like, cable bearing Bluetooth and VoIP, the telephone is no longer perceived as a rare resource. One of the characteristics of building a traditional network is that services need not be considered during construction, because the voice service is already embedded; thus, the service cannot be separated from the network. However, a network must have services (voice). This is the reason why traditional pricing can be calculated in each country according to the cost, regulation policy, ROI, market size, settlement between networks and user affordable value. Hence, a theoretical model of return on cost can be constructed.

As fixed line revenue is declining and increases in revenue from mobile voice service generally take three years longer than it does for a fixed-line service, a question automatically rises: why does the revenue derived from a mobile network not increase in proportion to an increase in network traffic, say 20 to 40%? If things are as they should be, the more traffic the pipeline provides, then there will likewise be an exponential rise in revenue as well. So, why is this not the case with mobile voice services? There must be some fundamental problem at work here that is not readily apparent. Presently, almost all operators are looking for some new kind of high profit services. Yet what kind of service should be considered as a high profit service? This is a question that is not easily answered at the moment. Providing the means to place an ordinary phone call cannot be considered as a high profit service. So, what will constitute the value and profitability of a new type service? If such a question is not able to be answered, then who can afford an annual negative 3-4% growth rate in the telecom industry? After most services become non-voice services, with no payment going to the pipeline, I think the question of how to create a model, which connects traffic in the pipeline, to the value of profitable services, should be considered in the new telecom economy.

As for the network, it will basically be divided into two layers: The first is CN (Communication Network) layer, and the second is the SN (Service Network) layer. The majority of new economic principles will come from the field of the SN (As is shown in Fig.1). In Europe, the following two aspects will be combined: operators will bear a kind of social responsibility - applying information technology to communication and similar type enterprises, which would include transferring social responsibilities like, energy conservation to those companies that comprise the telecom industry － which is already becoming a national policy. Now transformation has been enhanced to include another layer. Previously, in order to make things clearer, assessing such things as billing, or the ROI forecast were calculated according to the investment cost of tangible assets, while more knowledge-based or capital-based investments, which would include such things as: the investment in IT construction, service capability, application environment, middleware in software, along with other servers, were inadequately assessed. Therefore, it might be easier to calculate them separately.

Some traditional pricing methods are already in place in regards to the optical cable, switch and bearer equipment. There is a well-known story in reference to this. Originally, all investment went into accomplishing the same goal or purpose － enabling a phone call to be made. The price of a call was determined by means of the network investment cost and the marginal cost. After the price was set, the annual growth in the market would be the reward for the initial investment. However, when service products and the Internet began to flourish, they were replaced by the "bit" of 0101. Previously, the cost for 64K was high and was reduced by half, every 18 months, according to Moore's law. With a reduction in price, many new services began to emerge, especially non-voice services. At the same time, traditional voice service began to decline, with a reduction in bearer cost. Since the invention of the data network, virtually all transmission costs have been decreasing. As a result, the traffic of non-voice services has more than doubled. As a consequence of an increase in transmission overhead, non-voice service traffic exceeded that of voice traffic in the year 2000. However, its revenue was out of proportion with the increase in traffic. The construction level of a non-voice network is more than that for a PSTN (as Fig.1 shows). So, currently, much investment has no clear IOR model. That is to say, there would be no perfect profit model if broadband and 3G services were suddenly able to be commercialized. Thus, if things continue down this path, then a bubble will definitely occur in regards to network construction. Naturally, many in-depth issues and thoughts will surface if this becomes the case.

If one day all services should merge to form one giant service, or if the price was able to be completely calculated according to the bearer efficiency (China Telecom once tried to do this), then only 5% of the bandwidth would be able to cover the current voice traffic for the entire nation. Then, if the price for voice service could be calculated by way of bit profitability, then voice revenue would collapse as a result. Further, if the return on investment and cost could be calculated completely by data traffic, the benefit it would bring to the entire network would still not make it able to support future revenue. Hence, this is the challenge. How does one create and raise the profit brought over by the user consumption value (which may not be directly related to network nearer capability)? How can such value be linked to the cost and investment of the network? How can such a system be completed and regulated? How can reasonable investment be encouraged, and what is the justification for investment when more and more valuable services are beginning to emerge with each passing year? Thus, it would seem that there are many things that require further consideration on the part of the Telecom Economy Committee of late.

If one were to conduct a quantitative analysis of the traditional network, he or she may find that its value comes from selling minutes, bandwidth and/or the tangible asset it is operating on, while the value of a service network comes from selling capabilities such as: digital authentication, version watermark, fast downloading, searching, and storage resources. So, why has the mobile network done such a good job? It is because of its HLR, which helps sell service management capability and user management capability. The diversified portfolios of different services bring forth more and more services. Service itself is seen as a Rubik cube, and its application will have the effect of a Rubik cube. For instance, Chinese people do not like to use the voice mailbox. Although, almost everybody uses the mobile phone, some people are aware of a paradox when using the mobile phone, in that it sometimes bothers people's privacy, even though it provides convenience at the same time. Here is an example of a service portfolio: if the voice mailbox is combined with MMS, then, when a recipient receives a voice mail, he or she will directly see a photo of the caller, along with his or her MMS, without actually having entered the mailbox. In this way, SPs can make money from this type service, without annoying people, or offending the Chinese habit of not using voice mail. This represents a case of the so-called, Rubik cube effect. It is not only a kind of service, but also represents a mixed value like, "Super-Girl". Later on, a color ring back tone will not just be applied for listening, but also for viewing, then it will be possible for a service called, "color imaging", to be invented. Such a method serves as a model for creating new services. Traditional telecom economics cannot explain such issues as: the calculation of customer net value, marginal cost, and value allocation.


Customer Value and "Bit" Economics

The value source for a traditional application is essentially different than that for a new operation. When one speaks of the real source, using the voice mailbox service as an example, because it represents a good value model, then it changes people's perceptions. The value of consumption communication has no connection with the direct cost of a network. In fact, non-voice services have no relation with return on network investment. In the event that the 3G license should be released in China, all technologies such as: TD-SCDMA, WCDMA and CDMA2000, would provide call service. In such a scenario, the competition would not just be in terms of voice traffic －for low price competition (the first two phases of the BDG five phases) would finally become cost competition － but only in terms of service capability.If this were the case, then what would the sources of value be in terms of service capability? Speaking from the point of view of the service itself, its value comes from an operation enterprise's capability to provide services, customer's requirements for consumption and the mapping between requirements and the supply ability. Seeing it from a user's perspective, if he or she subscribed to both Chinavnet and Monternet, it would most likely also be the case that he or she could seamlessly use all the resources and services in both networks, without ever switching between them. Once he or she accessed the network, then he or she could also see his or her own homepage. Furthermore, when he or she is using some service, the issue with billing of services, regulation and settlement of cross-network services would be settled at the same time.

Currently, core competition is focused on the EPG (first user interface). The reason for this is that there is presently no coordinate system or standard equivalent for service inter-connectivity. To settle the account of services by way of the cost theory is actually very difficult to attain, largely because there is no explicit function relationship, between the value, and the cost of services. There is still a lot of controversy surrounding settlement between networks. Hence, relative justification in regards to this issue and theory development require further research, otherwise, it will influence the operation, competition and regulation of the future network. In settling the account of services, if the pricing of a certain service is not reasonable, there will be hidden competition and problems on another level when the service is being sold wholesale, or when the service interface begins to open up.

"Bit Paradox" is a highly significant phrase, which describes the complicated relationship that exists between, the bit bearer cost, processing cost, and service value in the future network. Please see "Bit Profitability", which is shown in Table 1. Here, each "bit" has an income value, which is listed from low to high. The income from a short message is quite high, over 1000 USD a Megabit, while that generated from a movie service is pretty low, probably several cents for several Megabits. Data services need a network to support them. Thus, if the entire network just had movies, its revenue would be terribly low. However, if the network does not provide movie services (do not use media stream), then the network cannot be rightfully called a broadband digital media network. Here arises the problem of the mismatch between, Bit Profitability, and the operation service value. The cost of each bit decreases according to Moore's law. Due to the reduction in cost, more and more services will begin to open up. Hence, more newly formed services will also be responsible for creating new costs for more service providers and new marginal costs for more services. Therefore, the operation cost is actually increasing, so a more effective solution for lowering cost is required. Hence, another contradiction surfaces: when the cost of each bit and profitability is declining, the network CAPEX should also decrease accordingly, but with the swelling of the CAPEX in the IT platform, the operation cost (For the digital media network, the expansion of the network does not mean an increase in services or revenue.) will actually increase, instead of decrease, to offset the loss of profitability. All these OPEXs come from the management of the many SPs and CPs, as well as from cross service provisioning. This means that the OPEX of each bit is increasing, and vise versa. That is to say, an increase of CAPEX for each bit, at the same time means a decrease of OPEX for each bit.

There is another law that explains it perhaps even better: The CAPEX of a bit is in reverse rate to the OPEX of a bit. That is to say, based on traditional investment theory, a reduction in CAPEX investment should be offset by the OPEX. The relationship between these two aspects, in terms of the total cost, is a subtle one. Therefore, investment in the digital media network requires a continuous reduction in bearer CAPEX, in direct proportion to annual bit profitability, which is in agreement with the natural trend. Just like in Moore's law, the income is decreasing, year by year. When the CAPEX cannot drop any more, it will lead to a loss in income. This will push you to take another step: increasing the CAPEX of the service network, while lowering the service OPEX. Indeed, there is a sophisticated curve that explains the relationship in the digital network, between bearer CAPEX and bearer OPEX, and the link between CAPEX and OPEX in the service network. In the U.S.A., Europe and in Japan, a lot of research has been conducted in this field, and the theories surrounding it are direly needed for guidance in regulation, service innovation and operation transformation, and are likewise required to direct operators in terms of significant investment in these areas.

A good example of this is shown in Table 1. Here, a comparison of profitability is graphically illustrated (out proportion to investment). It shows that when users download color ring back tone by way of short message, it will be charged 13,000 USD a Megabit, while the point-to-point short message charge is only 2,500 USD per Megabit. After 3G is launched, viewing a small screen will only cost the user ten or twenty cents. So, what does this mean? When you launch 3G services, a great deal of interactive media stream, or video information is also there, which brings down your profitability. But it cannot rightfully be called a 3G service, without such services. Hence, if you want to be a 3G service provider, in order to just provide this much service, current bandwidth and resources will not be enough; you will need to add more bandwidth and resources. So, during operation, issues concerning CAPEX, OPEX, profitability and business development, should also be considered in a more comprehensive sense. At the time when you decide to launch 3G services, where the legacy is 2G services, which use CDMA or GSM, you will find that ROI will become an issue, which further describes the concept of, "Bit Paradox". The concept of adopting bit profitability, new network architecture and cost analysis methodology, in order to deal with the research on how to invest and make a profit in a new network economy is called, "Bit Economics". We have had some research activities relating to it such as, the new Ramsey-Boiteux pricing, and the 3rd generation profit model. Generally speaking, such research is still not adequate enough to sufficiently guide future network construction and investment. We hope that economics and technical experts working in the telecom industry and Internet industry will conduct more research in this area.


OPEX and CAPEX in NGSN (Next Generation Service Network)

Investors in the OPEX of the service network are not just operators. For example, they would also include such entities as: MSN, Google, Yahoo, AOL, Sina, and Sohu. When these people come to invest, certainly there will be a common cost, which is commonly referred to as, cross cost. So, who should pay for this cross cost? We have engaged in some discussions with these kinds of operators: when there are thousands of SPs and CPs which need to be managed, the common cost of SP/CP management is far beyond the input in the platform. In this type of situation, its OPEX will be sky high. However, without the SP/CPs, there will not be any more revenue. When the revenue is climbing, the operation cost will increase by way of another index. This is a very complicated and difficult issue to resolve, which is causing a cost bottleneck to form, in terms of service and software operation. Thus, macroscopic regulation methodology and technology in regards to new application service are direly needed. This also creates a positioning problem for service operators, in terms of what to do, or not to do. In respect to the NGSN investment model, it will become more of a core issue in future network construction.

• Pricing for Customer Services
  
  In the future, there will be many kinds of services, and among them, you will also see the same type of service being offered to users like, short message, which will be provided by hundreds of providers. At that point there should also be a pricing principle(s) in place, and which one to use as an alternative to another, will be made available to providers, as well as different factors such as, discounts and preferential terms. In regards to future services, millions of customer groups (Each of these millions of customer groups is called, small customer groups. Customers using a phone call are referred to as, public customers. ) will face millions of services. People who are playing games, sending emails and playing with EDI (Electronic Data Interchange), respectively, will all belong to different small customer groups. Each service, in-and-of-itself will have its own requested value. It is not a matter of how much bandwidth the service uses. Each group of customers will have its own requirements in terms of services, and thus, will pay for the expenses of running and operating of such services. If they are charged by means of a flat rate take monthly as an example, when the N (the number of people in the customer group) becomes large enough, you will find that while these users were willing to pay for the monthly service when it was just beginning, however, they will be unwilling to pay for the marginal cost of having such as service, when service combination makes its usable scope smaller. Hence, you can see that the monthly-fee-mode is a concept that actually limits industry development, or it is only applicable in the early phase of service development.
  
  The pricing of service network capability is a kind market behavior, which ultimately determines the positioning of the network. Huawei is currently conducting research into this area (We call it, the 3rd generation profit mode). When millions of users, face millions of services, how does one quickly search for the desired target? A series of new capabilities and technologies such as, P2P, are very helpful for positioning a new network. In the society of consumption communication (commodity), when a telecom service is under commercial registration, its service classification should be clearly defined. Each service should have a clearly stated scope of pricing, along with the corresponding "transaction tax rate". During the course of a real transaction, the principle that buyers are responsible for clearly defining the payment mechanism, by quality, should be observed, and the price framework guidance for the operation of the market should be addressed as well. During an ordinary transaction, a relative “transaction tax rate” will also be paid. This is understood to be an associated network service platform capability. Once this sequence occurs, then the network can obtain more SP/CPs and provide automatic pricing and possess dealing capability under a regulatory circumstance.
  
  Based on such a model, as well as on the following conditions that network operation be divided into a bearer network operation and service network operation, then the OPEX and CAPEX of the bearer network operation will be clearly understood. However, the OPEX and CAPEX of service network operation still needs further study. Currently, the bearer network operation is slowing being transformed into the service network operation. After restudying the operation cost, operation pricing and customer value, operators will be led in a new direction: operators will find a new operation mode. Some virtual network operators (partners in the value chain) will begin to appear at this stage. Virtual network operation is an inevitable phase in telecom history, and it is also a part of the natural law of consumption. However, the natural law cannot be broken. It will come into existence, whether you agree with it, or not. It is very much like the initial phase of the Internet, which utilized a specific type of architecture. How to reasonably allocate the value that users have paid, and the surplus value to all the parts of value chain is, I think, a core issue that needs to be discussed in the telecom industry. Why is VNO so successful? Or, better yet, why is VNO so successful in countries where the information industry is more developed? Take Virgin as an example, you now see it all over Europe, but originally Virgin was only connected to the tape distribution industry, instead of VNO. However, once it had established its own brand, customer segmentation and special channels, it was also able to create a new model with network operators.
  
  Seeing it from a mobile value chain perspective, one observes that it has tens of parts. Some of them can adopt a cooperation model or distribution operation model, to some extent. Today, there are more than twenty foreign operators that have entered the Chinese market in some way or other. After once entering Chinese market, they almost always use the same procedure for searching for operators that will adopt a certain service from them and then distribute it throughout China, and in return they will both share the resources internationally. Data in connection with the total input in construction of all European mobile operators reveals that each subscriber invests on average 440 to 660 Euros toward construction costs, which is a rather large amount. Over 20% is invested by VNO, so that the investment can quickly go the proper place, with some being paid by construction CAPEX. After more and more VNOs enter the chain, the OPEX will then begin to decrease. It is similar to running a company limited by shares. Seeing it from the angle of mobile virtual network operation, one finds that as the number of subscriber's increases, together with the service revenue, that the cost relative to service income begins to decrease.
  
  
• Regulation and Telecom Public Service Principle
  
  Resale counts by volume: the larger the volume, the lower the resale cost. The same principle applies to operators. In VNO, the concept of volume has changed a little with the emergence of non-voice service operation. The value of most services is not necessarily counted by volume. There are many things that can be done in the bearer network and service network to balance service settlement or the borne settlement. However, if the cost is not closely met, many loopholes in cost will be utilized to meet this goal. For instance, there are many Internet service providers and infrastructure service determines (traditional telecom equipment provider) who will provide mixed services. However, there is no consistent criterion, method or measurement about how to evaluate user value and settlement, which results in value divergence and disorder.
  
  In reference to charging SP/CPs, they can be seen as pure service operators. If the previous payment was unreasonable: then they only pay for the bearer cost, by way of a monthly-fee mode, when they are using both bearer and service capability. Using the method of digital authentication, User profile management, storage space occupying, and a service tag to control charging, will force SP/CPs to pay bearer operators and to more closely follow the above mentioned customer pricing principle.
  
  After the transformation process in complete, there will be a lot of economic issues that will begin to surface in relation to non-voice services such as, how does one properly assess incumbent operators and in what industrial environment should they be placed. As to services, some are pure service providers. There are four models of transformation. Traditional operators will have four directions to go in after transformation. What will determine the value for these new service areas? For the phone call user, when he or she is calling, no matter who he or she is, whether it is a six-year-old boy, or a sixty-year-old woman, he or she will use the same mode to make a call. But now things have changed. There are now different small customer groups with various consumption behaviors, and the number of people in each customer group will not be the same. Their value cannot be described as easily as in Matecalf's law. There is also a concept of VNO. Each group will have their own brands. Suning and GOME may be the VNO in the future. However, this type of thing is not predictable. It will also be able to provide additional services like, the chain store, ctrip.net, and the airlines, and besides being able to provide a travel service, they will be able to exchange information with each other and obtain a cross value from it. Another alternative is to bundle billing, using a bank card. Previously, they were not able to bear bank services, or handle a large payment amount. However, if they form a partnership with banks, then the banks will deal with the payment, and the mobile services can finally be linked with payment. Research concerning small customer groups and total service operation will involve a lot of essential issues that are central to traditional telecom economics.
  
  As for public communication, it is reasonable for it is to bear services, but it is a little farfetched to make the investment in Internet and non-voice operation a public enterprise. As for consumption communication, which is found in the form of, short message, MMS, color ring back tone, and network poll, they are commonly viewed as commodities. However, controversy will ensue if the government intends at some point in the future to make services like these, public services. As for all global public service, the government will set the appropriate regulation and incentive policy, by way of a tax added value, to support public infrastructure construction. It seems that there is currently no similar policy for consumption communication (non-voice services).
  
  After the telecom industry completes its transformation, public service will become a social responsibility for operators. At the same time, they will be able to provide more non-public services. This is actually a kind of competition ability, and can be seen as another way to make a profit. A balance point between them has been found in many countries. A good regulation policy will encourage more services to emerge and create a more abundant market, where operators' can exchange their ideas more enthusiastically.
  
  And finally, let us look briefly at the service network. As a matter of fact, investment in the service network will be huge, similar to our legacy bearer network, and will include an infrastructure and IT facilities for the service network. The built-in capabilities will allow it to make money, and would include such aspects as, asset depreciation and cost. However, they would be different, in that they will have no direct relationship with service revenue. Generally speaking, many issues like the ones discussed above will pose great challenges to traditional telecom economy, especially in terms of investment in the service network. For operators and equipment providers, a lot of questions will require the services of industry experts and subject matter experts, in order to provide new and clear answers. It is not too much to say, that we expect that there will be a considerable amount of innovation and proactive insights that will emerge in this field, which will no doubt aid in giving the industry more explicit theoretical models that will help guide them in the future.