Tachikawa: staying one generation ahead

Published on:

NTT DoCoMo is one of the world's largest wireless operators, with over 23 million subscribers in Japan. It has also been promoting the adoption of W-CDMA as the global third generation mobile standard. And it is now establishing committees to work on 4G. NTT DoCoMo's president Keiji Tachikawa talks to Basil Ballhatchet about the operator's long-term plans.

Deregulation of the Japanese telecoms industry started in 1985, when the Nippon Telegraph and Telephone Corporation (NTT) was transformed into a private company. After six years of protracted discussions between the ministries of finance and posts and telecommunications, it was decided to spin off the mobile operations from NTT.

The new company, NTT Ido Tshushin Kikaku Kabushikikaisha, was created in August 1991. The name was subsequently changed to NTT DoCoMo in April 1992 (with DoCoMo standing for "Do Communications over the Mobile Network"). In July 1993 NTT DoCoMo transferred sales activities in the mobile communications business to eight regional subsidiaries. The operator rapidly became the biggest mobile operator in the world. It established its first subsidiary in Europe in August 1998.

In October 1998 a 28.5% stake in NTT DoCoMo was sold in an initial public offering. The IPO was more than two times over-subscribed, raising $18.4 billion. Following the listing, NTT's holding fell to 67.1%.

Kate Lye, a telecoms equity analyst at Warburg Dillon Read, talks about DoCoMo's recent success and the impact of the new CDMA offering from DDI and IDO: "In the last year, it has been an excellent performance. They increased their cellular subscriber base by 30% in the past financial year. April 1999 has been a point of change, with the introduction of cdmaOne on a nation-wide basis from DDI and IDO. Sure enough, we have seen some impact on their market share. To provide an example, DoCoMo's market share in Tokyo fell to 48% of new subscribers. Their cumulative market share has been 68% of new subscribers over the past year. So cdmaOne is certainly having some impact.

"Having said that, particularly in the case of DoCoMo, I am not sure that market share is necessarily the main issue. I think that profitability is what we want to see. I think that DoCoMo will take steps: they are not going to jeopardize profitability, but they are also going to pretty well maintain their market share."

Lye explains how DoCoMo may evolve its strategy, to counter the increased competition from DDI and IDO: "I think that they will try to maintain their high-quality, high-usage customers and that they will continue to provide very good after-market services. Secondly, they will try to win on services and products, particularly data and their i-mode service. I think this will come in above the cdmaOne services. Thirdly, I think they will take somewhat of a counter-active action against cdmaOne. They are quite specifically taking three months to watch what happens with cdmaOne. If they seem to be losing substantial market share, they would then look at new pricing strategies, different pricing plans, perhaps price cuts - new services - expanding i-mode."

NTT DoCoMo's i-mode service enables subscribers to access Internet services through their mobile phones. Users gain access to a number of on-line services, including interactive services, such as mobile banking, news and stock updates and ticket reservations.

As i-mode is based on packet data transmission technology, users only have to pay when they actually retrieve information, rather than the time that they are on-line. In addition i-mode customers can use to the service to send and receive e-mails through computers, PDAs (personal digital assistants) and other i-mode phones. The e-mails are displayed as soon as they arrive. Launched in February 1999, 220,000 customers had already subscribed to the service by May 23. The operator expects to have 4 million subscribers by the end of the year.

NTT DoCoMo has also carried out R&D on advanced mobile communications technologies, promoting its own wideband code division multiple access system (W-CDMA) as the international third generation standard. At a meeting in June 1999 it appeared that an agreement on a group of standards, based on CDMA, was supported by operators serving 75% of the world's customers. However, it had still not been signed by some US wireless operators. NTT DoCoMo seems likely to be one of the first operators to adopt a 3G system in 2001 and is even establishing committees to work on developing a 4G standard.

In an exclusive interview with Global Telecoms Business, NTT DoCoMo's president Keiji Tachikawa talks about third generation, the i-mode service, the development of wireless data applications with Microsoft and other revenue drivers that will enable the operator to remain at the forefront of wireless communications.

It looks as if there may not be one universal 3G standard. How do you react to these developments? Does this imply that you will delay implementation of W-CDMA in Japan? Are your competitors, DDI and IDO, stealing a march on NTT DoCoMo, as your 3G services will only be available in 2001? How much is the CDMA network costing to build?

Well let me start with the next generation standard issue. It seemed at one time that there would be some trouble because of the Ericsson/Qualcomm battle. However, that problem has already been resolved. The ITU's standardization initiatives are making progress and standardization is almost proceeding according to plan. However, the original goal of the next generation system, which was to achieve a perfectly single standard, is not really going to be realized, as many countries oppose this idea, in particular the US. So multiple standard technologies may well co-exist in the market.

We have also changed our position on this matter. We believe that the existence of multiple technologies that compete with each other might be better and beneficial in terms of joint technological development rather than having one single standard. If you only had one technology, the technological life time is only about 20 years. Therefore, instead of having only one unified standard and people becoming more conservative to preserve that standard, it might be better to have more technologies.

In Japan another operator other than DoCoMo might adopt wideband CDMA for the next generation system. As you mentioned, DDI and IDO have not yet decided on their preferred next generation technology. I expect that they will use the US technology. So different technologies would co-exist in the Japanese market.

Admittedly starting this spring DDI and IDO launched a 21/2G technology nation-wide, cdmaOne. This technology seems to have been well received in the market, as it has new features compared to 2G. So that technology represents a threat to DoCoMo, but that is an aspect of competition. We are responding to this threat by resolving the problems in the 2G network. And 3G systems will take care of those issues that cannot be resolved through these improvements.

So when it comes to the high-speed transmission capabilities, such as 64Kbps or 128Kbps, those rates cannot be provided to our customers efficiently in the current network. So we believe that it is better to migrate to 3G systems in those cases. We are witnessing increasing demand for data communications in the market in Japan. To meet these demands, we would like to introduce the 3G system in spring 2001, as scheduled. The necessary preparations are now under way.

In April we completed the selection of joint development partners for the equipment to be used for third generation systems. We are assuming that all the details of the specifications will be finalized in December this year. So we would like to start the development and production phase accordingly. We are now planning to start network deployment and construction around the middle of year 2000, but we have not yet decided how large the network will be in the initial stage. That is because we have still not judged demand for this network. It is supposed to cater for high-speed data capabilities - we have not yet finalized our estimation of actual demand.

Therefore the total amount of investment has not been determined. But we would like to achieve a lower cost compared to second generation systems. So naturally if you compare voice communications, specifically this means that the 3G system should cost less than the 2G system.

If I can take you up on two of the points that you made. Firstly on improvements to 2G technology. A number of operators have referred to the benefits of GSM 21/2, implying a phase between 2G and 3G systems. Some people have claimed that packet GSM could be a potential replacement to 3G. It has been stated that these technologies provide ample data capacity. How do you view such announcements?

Compared to our PDC system, GSM has the potential to offer slightly higher transmission capabilities. The introduction of packet technology may achieve a bit rate of 115Kbt/second. I have heard that some European operators are investigating possibilities in this area. I think that it might be a good idea to absorb immediate market demand.

However, in Japan we have already introduced a packet data network, using the existing PDC system, which offers a bit rate of 28.8Kbps. This was introduced two years ago. But unfortunately it is not really efficient to try to achieve a higher bit rate than 28.8.

We would expect to achieve higher rates through the 3G system. But in any event even with GSM it is not feasible to achieve a higher bit rate than 384Kbps. So a 3G system will be required at some time in the future.

In the end, it is a matter of the strategy that you decide to take: would you like to extend 2G service as much as possible or would you prefer to select 3G and introduce it as soon as possible to try and deliver a full coverage of service with 3G at an early stage?

There is another constraint in Japan: as most operators are making good use of 2G spectrum, in order to provide high-speed data capabilities it might be better to utilize a new frequency band and achieve a new service by obtaining new spectrum from the operators.

Certain operators and suppliers have been demanding a phased approach to 3G capabilities, involving interim enhancements to US-TDMA, CDMA/IS-95 and GSM. Could attempts to hold back 3G systems and the increasing competitive climate, coupled with the consequent need for further R&D expenditure, adversely affect NTT DoCoMo's margins?

I don't agree. The 3G concept was raised almost 15 years ago in 1985. So we have spent 15 years on the R&D into this concept. It must materialize over the next few years. If we don't move forward quickly, we will be surpassed by 4G. And we have already started the R&D committees for 4G. We are targeting 2010 for 4G.

Could you tell our readers about the W-CDMA handsets that will be available when the network is up and running? What added capabilities will these handsets have?

The key factors offered by 3G terminals are high-speed capabilities and the wideband capabilities. So we believe that 3G terminals should take full advantage of the high-speed and wider band capabilities. At a meeting of our joint development department the other day, we focused on four types of terminals to be jointly developed and supplied.

The first is a hand-held type, which could be used as a TV phone as well. This is a portable phone which would carry both voice and data capabilities of up to about 128Kbps. It would also be used for image transmission.

The second is a car navigation terminal: it will have a larger screen and will be mounted on a car. We are also considering the possibility of mobile phones equipped on top of a personal computer, with colour moving pictures transmitted through the Internet.

But it is easier to assume that these personal computers would be used in a stationary environment rather than on the move. So this would have a transmission capability of up to 2Mbps.

And the fourth type of terminal is more suited for inter-computer communications or machine-machine communications, which will not handle voice at all. While there are others, we would consider first these different types of terminals which will represent a striking difference from 2G terminals.

Which suppliers are you using for the handsets?

There is an extensive list of suppliers, as we have classified them for each type. There are 15 vendors. The foreign vendors include Nokia, Ericsson and Motorola.

Why did you select Lucent Technologies to supply the Base Transceiver Stations (BTS) and the Radio Network Controller/Multimedia Processing Equipment (RNC/MPE)?

Actually our procurement process dates back to the initial process for the experimental system. We signed contracts with the vendors that passed our exams, looking at their track record in the experimental system and that of the other vendors for the proposal of the commercial system.

In short we selected vendors that can meet our specification and timing restraints. We also look at costs.

Could you tell us about the "Doccimo" service? How important is it to combine PHS and cellular in one phone? What do you perceive to be the benefits of PHS services?

So far people in Japan tend to think of telecoms services quite independently of each other. But in my view there are fashions and trends: a specific service prospers and then declines. There could be a supplementary relationship between services. In my opinion PHS and cellular can provide very close services to each other, centred on one service. So it is a very close system offering mainly voice services.

There are some specific features to both: the PDC cellular system can handle high-speed motions, such as the speed as you move by car. On the other hand the PHS service can only cater to a walking speed. In contrast the bandwidth is different for both. This means that PDC can only offer a bit rate of 9.6Kbps, while PHS can offer up to 64Kbps.

So you cannot forget about these differences. To achieve the greatest benefit, it is best to combine them. To attain this goal, we developed the composite terminal, Doccimo. Naturally the advances in production technology achieved by the manufacturer contributed to the materialization of this product. Doccimo weighs less than 90 grams.

I believe that this would offer a significant benefit to users, as they will be able to select the mode that they would prefer to use, depending on their own environment and usage.

After we introduce the 3G system in 2001, we believe that it would be best for us to combine the benefits of these systems and pass them on to our customers. I am referring to W-CDMA, PDC, PHS and pagers. Even pagers can be used for an information delivery network, with a data rate of 6.4Kbps.

How are you going to try and boost revenues for your PHS services?

Actually an increasing number of our subscribers appreciate the data rate of 64Kbpsoffered by PHS. We have witnessed a net growth in PHS subscribers since April 1999. So we anticipate growth in subscriber levels and consequently in revenues.

More interestingly we now plan to develop a location and detection service, based on this technology. So at a half or a third of the size of a cassette tape box, this terminal will have no voice or communications capabilities, but will be dedicated for location/detection. This new type of terminal can be used for small children and elderly people and also for pets. So we would like to secure significant PHS revenue from this type of new service.

Why do you think that the Ministry of Posts and Telecommunications recently over-ruled DoCoMo's discount package on cellular-PHS bundle contracts?

This was not really a policy decision of the state authorities. They were responding to complaints raised by our competitors: as they cannot provide the same service - as they do not have both - they wanted to delay the introduction of this service.

It is true that the ministry asked us to review the tariff structure, as there were some unreasonable points in that structure. They decided to adopt a wait and see approach, since they do not believe that there is an inherent problem in the bundling.

So the basic element of that ruling can be summed up as follows: the elements that existed here were not included in the other plan. To consider user benefits, we decided to have an equal level in the menu. In other words the discount rate is much higher than we had originally planned. So they helped us in the end.

A number of analysts believed that this ruling might lead DoCoMo to exit from the PHS business. They also believe that withdrawal from PHS would be a good step to make, as this would rule out any potential dilution to DoCoMo's earnings. What do you think?

There are such analysts in the market. But we believe that if we have a comprehensive range of products, we can diversify the services that we offer our customers. In addition, one of the characteristics of DoCoMo is that we are offering a comprehensive range of services, including maritime telephony and in-flight service. We also have satellite communications.

So instead of splitting off the PHS business, we believe that it is better to offer comprehensive services and combine different elements. In our opinion it would be more beneficial to offer our users a more composite range of services.

I realize that some analysts out there are saying that paging is a loss-making business and that therefore DoCoMo should stop this service. But rather than dumping this system, DoCoMo believes that pagers have a good potential for multicasting information. So this will now become the core of the future pager service.

How do you view the market for wireless data applications in Japan? Which new services will you be able to offer customers in 2001? Do you think that applications such as Internet access and televised phone conversations will boost growth in subscriber numbers?

We are already responding to requirements for Internet access and e-mail services with our network. In terms of e-mail, we are offering a short messaging service and ten yen ($0.08) e-mail service.

We have both a circuit-switched network and packet-switched network for Internet access. Unfortunately even if we already have those in place, data communications account for only 1% of the total traffic.

In contrast in the PHS network data traffic already accounts for 25% of the total. We signed contracts with the vendors that passed our exams. With the PDC system we aim to increase data traffic to 10% by 2001. At about the same time we are planning to introduce the third generation wideband CDMA network. So among that 10% traffic those who opt for high-speed data transmission capabilities will be transferred to the wideband CDMA network.

And then further down the road - in 2010 - I would forecast, although this is not definite, that data traffic would overwhelm voice traffic. I predict that there will be 80 million cellular subscribers by 2010 in Japan, which is actually double the level in 1999. In other words these 80 million subscribers will represent an aggregation of the users subscribing to multiple services (for example three services for one person).

So let us say that the user would carry in his pocket a handset with a TV. In his car he would have another terminal for car navigation. The user would have built-in transmission equipment in the PC. With this alone you already have three subscriptions. Only one of these three terminals is for voice, the others are for data traffic. So this explains why data traffic would overwhelm voice traffic in the future. This would mean a huge number of terminals out there in the market.

What percentage of the market with you be able to retain in 2010?

I don't know. I cannot forecast. We have 57% now. If we manage to maintain the status quo, we will retain this share. I want DoCoMo to remain the leading company.

What do you perceive to be actual demand for mobile computing in Japan? Can you explain how the tariffing system could enable consumers to have a free 24-hour connection to the Internet?

The traffic for mobile computing amounts to about 1% of the total traffic. But then you need to consider the number of terminals sold: that would include the data cards, the PDAs and composite PDAs. The total number of minutes sold for mobile computing already exceeds the one million mark.

In February this year we launched an interesting new service for our customers, the i-mode service, which is a hybrid of the packet network and the traditional PDC network.

So when i-mode is connected with the packet network, it would be used for the transfer of information, especially data communications. The terminal is constantly connected to the Internet, but the customer will not have to pay any tariff, if no information is transferred..

However, in the radio environment it is never possible for a constant connection, since that would consume the limited spectrum resource. So only the packet solution will virtually enable that to happen. Therefore we are utilizing packet. But from the consumer's viewpoint it looks as if the consumers are always connected to the network and that they can always contact the Internet, whenever they want.

Chess provides a good example. If you are playing a game of chess with a remote partner, utilize a circuit-switched network and the game took five hours, you have to pay a tariff for a five-hour phone call. However, when you use a packet-switched network to play this chess game, you will only pay for the moves that you make. So the information transacted is very small.

This is one solution that we would like to use more, as we respond to demand for Internet access. That is the basic concept of next-generation wideband CDMA. So you can say that wideband CDMA will be fully used for packet connections.

You mentioned i-mode: you already have 220,000 subscribers after three months. How many subscribers do you expect to have by 2001?

Our first target is to achieve 4 million subscribers by the end of 1999. I believe that we would like to have 50% of all our customers connected to i-mode within two or three years. At this time i-mode will have some other features. In other words the terminals to be released from next year onwards will already have standard i-mode equipment. The phone that you buy now is equipped with i-mode.

Could you tell us about your relationship with Microsoft to develop wireless data applications? How do you plan to use Sun's Java in cell phones? What does NTT DoCoMo gain from these strategic relationships?

When you consider computer communications on the move, you always have to think about the limited resources of the radio spectrum. But if you look at the traditional protocols, they are basically developed, envisioning a fixed environment. So we believe that the PCPIP protocol is not really suitable for the radio interface.

In addition it is very unlikely for us and not desirable for us to install that heavy Windows 98 operating system on the terminal. There should be an operating system and protocol that is more suitable for the mobile terminal.

So we have just embarked on a mutual development with Microsoft, listing our requirements. But at the same time we are not totally reliant on Microsoft. We are looking at Symbian as well, as they are developing an operating system. So we are involved in joint developments with Symbian. Going forward the users will in the end choose the one they prefer.

Our relationship with Sun is related more to the protocol aspect. The market has not still decided which is the best protocol for the radio environment. We are using html in the current version of i-mode. But we know that WAP (wireless application protocol) is there. So going forward the market will decide which is the best protocol.

One unique feature of i-mode is a good big-sized screen. It also weighs about the same as ordinary phones.

Which new telecoms services are you providing for the increasing number of old people in Japan? How are you attracting more female consumers?

We have actually taken a number of measures to incorporate more female customers on our network. And we have been successful in this area. For Japanese at least, women tend to talk much more to each other than men. The vast majority of telephone calls on the fixed-line network are due to women. And this is also true for increases in the mobile business.

This applies only to the younger generation and not for the aged. But that younger generation will grow and at some time in the future everybody in Japan will perhaps love chatting on the phone. This means that the president of NTT DoCoMo 20 years from now will have a very easy time with everything.

At this time women aged over 40 are not really interested in having a long chat over the phone. Men over 60 are not really interested in mobile phones. So these constitute some of the targets for DoCoMo. The immediate target is the one who appreciates the convenience. The next target relates to computer communications.

So we are targeting the younger generation, both men and women. As this generation is brought up on and very familiar with games. we decided to target them. Demand for computer communications using a cell phone is mostly driven by men and women in their 20s and 30s.

Initially women in their 20s lagged behind men in terms of e-mail use. Therefore our company developed a very inexpensive and easy to use keyboard, which can easily be connected to a mobile phone, in order to develop demand in this sector. We achieved record sales of 300,000 units in only a year. The keyboard terminal was developed by female employees of DoCoMo.

An estimated 88% of the population in Japan and 57% of 15-19 year olds have a mobile phone. Do you agree that mobile phones seem to be replacing other leisure goods in Japan? Have you noticed any reaction from the leisure industries?

Some industries are indeed complaining that mobile phones have taken away their business. I would treat this complaint as follows. If you look at the ARPU, it works out at only 9,000 yen ($74.28) on average. Naturally the total 9,000 yen does not all go to DoCoMo.

But in any event this monthly ARPU indicates that an insubstantial proportion of disposable income is spent on mobile phones. These young people have more money to spend. They are spending money on other goods as well.

It might also be possible to say that, if you consider the times when they do not have a mobile phone, they have to spend 9,000 yen on average anyway. Some people claim that comic books are not selling as well as in the past or that the fast food industries are achieving less revenues than before. I would say that the young generation previously read too many comic books.

How long do DoCoMo subscribers keep a handset, before switching to new smaller models? How much does it cost DoCoMo to subsidize the frequent switch-over from handset to handset of existing subscribers?

Our company puts a new terminal in the market once a year. That is a reasonable period to cater for technological innovations. That is part of our competition policy. But I do not want users to replace the terminals too frequently, because every time they do we have to spend some sales commission on each terminal or cross-subsidize customers.

It is also true that things are becoming controversial. That is because we have 25 million customers on our network. Every customer replaces their terminal once every three years. This means that every year 8 million new terminals are issued. If we have to pay so much commission to cross-subsidize users a lot, this will affect our business. We do not encourage so many replacements.

IDO is gaining a number of new subscribers through a series of free minute pricing plans. Will you respond with something similar?

We have developed a similar service already. On July 1 we will offer a new service with more hours of free minutes.

What levels of churn has NTT DoCoMo been experiencing? How does NTT DoCoMo plan to retain customers? With such a high overall market share, is it a case of boosting ARPU rather than market percentage? How will you continue to improve cost efficiencies at a faster rate than ARPU declines?

Churn is a universal problem in the cellular sector. One thing that we have always done at DoCoMo is to ensure that the churn rate at DoCoMo has been very low. We would like to endeavour further to keep the churn rates lower. But it is very difficult to keep churn low in such a competitive market. By offering attractive terminals and attractive services for users, we intend to keep these levels low.

I would say that the decline in ARPU is partially attributable to the reduction in tariffs. We have observed another interesting trend recently: the new subscriber is not necessarily accounting for less traffic. When a high-school boy graduates and becomes a university student, this new student will use mobile phones more frequently than a 30-year old or 40-year old man.

Or rather these kids might use the phones for longer hours at midnight. So it can also be said that while the minutes of usage may grow, ARPU will not grow accordingly. If you use the phone at midnight, it is cheaper. So we analyse these things very carefully when we reduce our tariffs. In addition, with better cost efficiency in management of the company, we are trying to reduce tariff levels.

How are you going to balance the acquisition of new subscribers with network congestion in Tokyo? How much capacity do you have left? Will you be able to buy more spectrum?

We have always struggled to achieve the best spectrum efficiency. Thanks to these efforts we are ranked number one in the world in terms of the efficient use of spectrum. Basically if we have a shortage, we have to obtain the new spectrum.

In the past when we encountered this problem, we managed to obtain new spectrum. And that is the basic policy to meet with the congestion problems as we go forward. But it is also true that the Japanese government is not so easy to handle on this issue. Unless we can prove efficient use of spectrum, the government is not that ready to provide additional spectrum.

It seems likely that a government decision on the allocation of spectrum could be delayed. Is there a possibility that the launch of your W-CDMA service could be postponed?

No. I don't think that it will be delayed. The government position on the allocation of spectrum is that they would license more spectrum usage, once the ITU standard had been fixed. The ITU is moving on schedule. So I don't think that there will be any delay in spectrum allocation.

Do you believe that there will be widespread consolidation in the Japanese cellular industry over the next two-three years? Do you see international operators stepping up their presence in this market?

On July 1 NTT will undergo restructuring. To meet this change, the other competitors will have to restructure and re-organize. I would forecast that operators that provide only one single service may not survive the competition in the long term. That was the case for PHS. It is now becoming a problem for pagers.

And I don't think that such a service would last in the market without some integration. As I said at the outset, if you want to survive in the future, you have to consider user benefits by combining services and offering a diversified menu to users.

It has been reported recently that NTT will sell an additional stake in DoCoMo, reducing its stake from 67%. How do you view these rumours?

The president of NTT has announced that for the time being they would at least try to retain 50% or more. And that may be the minimum position for NTT, as it effects the transition to a new structure, with the holding company and management of the whole group. So 50% might be the minimum requirement. It is not up to me to decide whether to reduce the stake from 67-50%. That is up to NTT.

Currently, you are ranked second to SmarTone in terms of highest revenue/subscriber. SmarTone has high roaming revenues. Do you plan to link up with international operators to provide international roaming services?

Roaming is just about connecting your network with another network. This means that the revenue has to be split. So in that sense I don't think that roaming is more attractive than the provision of telephone services at home. So if you want to have a direct boost in revenue, it might be better to go abroad and provide directly services there.

International roaming is a very good approach to adopt to increase traffic. In that sense it is beneficial to both partners. DoCoMo has already established roaming arrangements with many parts of the world.

What are your hopes and ambitions for the company over the next two-three years? What trends do you see emerging in global wireless communications over this time frame?

For the coming two or three years, our objective is to implement smoothly the next generation system and expand data traffic to 10% of the total, so that we increase our share of the wireless market.

That would be the springboard for further goals in the future. If you look at the past five years, market growth was so steep. Now we are coming to a plateau. But if you do the right thing over the next two or three years and prepare a good springboard for the future, you will boost growth over the next 10-15 years.

Globally I think that everybody will be targeting the same areas. One important thing is that everybody should pursue the original target: users should be able to use their own phones wherever they are in the world. I believe that the other wireless operators are all anticipating the progress of mobile multi-media. And that is the source of revenue moving forward.