Has i-Mode Lost its Mojo?

Back to Contents of Issue: October 2001


Or is it just the costly recalls, damaging delays, decreasing growth, and increasing competition?

by Daniel Scuka

THE END OF SUMMER in Japan brings sultry, humid weather to much of this island archipelago, and city dwellers traditionally flee to the cooler countryside for Obon holidays to honor deceased ancestors. By late August, fleeing is probably what the overseers of NTT DoCoMo's i-mode felt like doing -- and not just because of the heat. Much of the recent news related to the world's No. 1 wireless information service has been bad, if not downright terrible, and while there's been little public acknowledgement of the situation, it's starting to look like i-mode is facing tough times ahead.

First, there's been a spate of expensive and embarrassing recalls of i-mode handsets -- of both the current 2.5G and the newer, still experimental, 3G models. Further, the i-mode network has been plagued by the continuing stench of spam, driving some customers to complain long and loudly. Just when that situation seemed to be under control, senior company executives revealed that the rollout of i-mode services in Europe, previously scheduled for this fall, would be delayed, with no new date set. While these events alone would be sufficient to cause gloom at the Sanno Park Tower, there's worse news, and it's related directly to i-mode's bottom line.

Specifically, a close look at DoCoMo's subscriber growth and average revenue per user numbers (ARPU is the gold standard by which all telcos are judged the world over) indicates trouble on the horizon. Existing subscribers are already using voice and data (i-mode) as much as they're going to, growth in new subscribers is falling off, and -- even if new subscribers were increasing -- the newbies don't use the network as much as the early adopters who have already signed up.

The carrier appears to be in the early stages of what's called the marginal subscriber effect, where it is faced with ongoing costs to acquire each additional subscriber and operate the network, but the newcomers don't add to usage (or revenues) like the earlier ones did. "All carriers eventually face diminishing returns due to the marginal subscriber effect," says Kieran Calder, senior telecommunications analyst at Indosuez W.I. Carr Securities in Tokyo. Voice is still the killer app for any mobile network, and voice ARPU -- steadily falling for a number of years -- is clearly already feeling the marginal subscriber effect. "DoCoMo is facing a period where growth will slow and returns will fall. Voice ARPU continues to fall by 10 percent per year," wrote Calder in a May 2001 research report. In fact, monthly voice ARPU has fallen from JPY15,930 in 1996 to JPY8,620 in 2000, and predictions for the July to September 2001 quarter range from a high of JPY7,160 (according to DoCoMo) to a low of JPY7,050 (according to a senior analyst at a European investment bank in Tokyo who requested anonymity). DoCoMo managers themselves expect voice ARPU to fall another 7.9 percent by the end of fiscal 2001 (year ending March 2002), and almost no one expects this trend to reverse.

This means that any growth in total ARPU has had to come from data, but that, too, has now entered a downturn in growth. The situation is so serious that some analysts, like Calder, are predicting that DoCoMo's November report for the July to September quarter will show a drop in total (voice plus data) ARPU, making the company unattractive to investors and hurting the share price. Calder forecasts data ARPU for September to be JPY2,068, down from a peak of JPY2,300 in the April to June 2001 quarter. With this downturn in data ARPU, the long slide in total ARPU has begun.

i-mode, You mode

There is, mind you, not much danger that DoCoMo will miss payroll anytime soon. The 900-pound gorilla of Japan's mobile industry is, by all current numbers, still doing extraordinarily well.

A July 31 report issued by the Telecommunication Carriers Association indicates DoCoMo is the domestic market leader for subscribers, with 37,784,000 overall, 26,085,000 of whom subscribe to i-mode (that's almost as big as the population of Canada, to put things into perspective). The July numbers represented a month-on-month increase of 1,096,000 i-mode subscribers (wow!). The importance of adding new users becomes apparent when one considers that each i-moder will generate an additional JPY1,420 per month in ARPU (due to data usage and site access fees) for FY2001, according to a company forecast released on August 10.

Further, DoCoMo has seen a number of new service deployments this year that have contributed to revenue and enhanced buzz -- something that may be more valuable than immediate revenue in the hypercompetitive Japan mobile market. Early this year, i-mode was the first to market with Java-based services, launching on January 26. Java permits small applications to be downloaded to the keitai handset, so that games, financial, or other interactive applications can execute right on the phone. DoCoMo's Java, under the i-Appli brand name, is proving highly popular, and as of August, there were some 120-plus sites offering Java downloads.

Competitors J-Phone and KDDI only came to the Java table in June, and their services were still in the early days as of press time. KDDI is also planning to launch a BREW service later this year. (BREW -- for Binary Runtime Environment -- is a technology developed by Qualcomm in the US that offers programming functionality similar to Java. Both allow mobile developers to deliver small, highly interactive applications -- including games, real-time stock quotes, and other business and entertainment services -- to wireless-enabled devices.)

Both J-Phone and KDDI, desperate to compete with the DoCoMo Java juggernaut, have announced that their services will permit larger Java downloads than DoCoMo, at 30 and 50KB for J-Phone and KDDI, respectively, compared to i-Appli's tiny 10KB. "This was done deliberately -- as a point of competitive differentiation," says Masako Harada, CEO of Kanagawa-based game software developer inis. While that may help make the non-DoCoMo Java services more attractive to game and other developers, DoCoMo has a huge lead. In August, there were already 5.5 million DoCoMo Java users, while J-Phone and KDDI had only just started their services.

DoCoMo also upgraded i-mode by introducing the new i-area service in July. The i-area system (accessible via the regular i-mode menu) automatically gathers and displays information concerning the caller's location and neighboring areas. The handset's position is determined based on cell location, and users can view a weather forecast, look up a map based on address or place name, search for a nearby restaurant, check on traffic congestion, or check hotel information. While no usage numbers are yet public, one company source claimed that i-area is "doing well."

And when it comes to buzz, nothing helps like launching a faintly goofy, ultra-kawaii (cute) consumer service that meshes with mobile, which is what DoCoMo's new c-mode vending machine service does. The c-mode system is basically a Coke vending machine transformed into a keitai-controllable soft drink-cum-multimedia service kiosk, and DoCoMo, Coca-Cola Japan, and Itochu worked for more than a year to develop the concept. c-mode was a tough sell even to i-mode's senior people (insiders relate that it took four attempts before i-mode's master, Takeshi Natsuno, would even consider the idea), but it may just work. The trial machines (in Odaiba and outside Shibuya's 109 building, among other areas) vend, obviously, Cokes, as well as ring tones, screen wallpaper, maps, merchant coupons, and local information. Users earn points, and pay for purchases through a debit system.

DoCoMo was also one of the world's first carriers to roll out 3G cellular services, at least on a trial basis. It began serving some 4,500 test subscribers in May. The company plans to open the network to new subscribers about the time you read this, and by the end of July, the W-CDMA system was already providing more or less reliable data downloads at 200-plus Kbps, faster than any other wireless network anywhere.

i-slowed But despite this apparent succession of i-mode triumphs, the company is facing a growing list of problems. The handset recalls mentioned earlier were an unmitigated PR and marketing disaster for DoCoMo and the manufacturers involved: Sony, Panasonic, and NEC (the latter for 3G handsets). While the public was placated with full exchanges and free upgrades, the DoCoMo-Sony inter-company relationship, for one, must have taken a hit. Sony announced in July that it would take a $95 million earnings charge related to recalling faulty handsets supplied to DoCoMo and other carriers.

The uncertainty related to the rollout of i-mode-like services at DoCoMo's European partners is also a significant setback. Privately, company officials admit it will take longer than they thought to get the European partners onboard with i-mode. That's due to technical difficulties with GPRS (General Packet Radio Service) upgrades to existing 2G GSM networks -- which are required to offer any sort of wireless information services based on the i-mode model, delays in obtaining new GPRS handsets, and the time needed to educate partners on how, precisely, to adapt the i-mode model to local markets. One move under consideration would see the dispatch of experienced i-mode planning personnel from Tokyo to the European partner companies to offer long-term, hands-on assistance.

Also, to counter the bad karma from all the spam, DoCoMo graciously announced in August that it would allow i-mode subscribers a credit equal to 400 packets of data each month on their accounts, in theory making the receipt of the spam free (although this move does nothing to reduce the spam itself). Seems fair enough, but remember that there are 26,085,000 account holders (as of August 17), and at JPY0.3 per packet, that translates into quite a pile of cash. DoCoMo may have deep pockets, but foregoing this much revenue has to hurt, and investment analysts are already discounting the company's earnings in their fiscal 2001 performance forecasts.

Packet Ping-Pong Analyzing packet usage is key to understanding DoCoMo's predicament, and Java has been one of the main drivers this year. Packets are why i-mode has become the cash cow for DoCoMo, and packet fees generate some 10 times more revenue than other non-voice mobile revenues (including i-mode content access commissions, now set at 9 percent). Non-Java packet usage per subscriber has been falling steadily, from 200 to 180 packets per day during the September 2000 to June 2001 period. Indosuez analyst Calder expects this trend to continue, and is forecasting an additional drop in non-Java packet usage of 5.4 percent to 151 packets per day (including the 400 freebies offered monthly to account for all the spam) by the end of September.

The Java packet situation is even drearier, with usage falling faster than Tokyo real estate values after a strong initial showing of some 500 packets per day. Calder expects Java packet usage to tumble to 348 per day (down an additional 6 percent) by September if first quarter usage trends continue in the second quarter (this drop also includes the 400 free packets).

Many industry watchers believe that Java could provide additional packet usage if new applications could be found. However, so far, most Java usage has been confined to games and some consumer-targeted services like stock quotes. There are few enterprise or business applications available, and in this, DoCoMo is no different than its competitors J-Phone and KDDI. "The big operators have the networks, but their mindset is consumer. That will last for a long time," says Punnamas Vichitkulwongsa, founder of Tokyo-based mobile enterprise application developer Arriya Solutions. Privately, DoCoMo business development staff are optimistic that the next generation of Java-capable handsets, the 504 series, will provide a fresh boost in Java (and hence packet) usage. One unconfirmed rumor is that the new model series will provide an even more robust Java environment than at present, with downloadable Java applications being able to access expanded onboard memory. This would allow, for example, email file attachments and more useful Java functions.

But unless the company can figure out how to exploit the business and enterprise market, any such a boost is likely to be temporary. "Sure, usage might go up," says Calder, referring to the rumored 504 capabilities, "but maybe not." As for the rumors of expanded Java environment, Nobuo Kawakami, CEO of Java game developer Dwango, says: "I don't think it will happen -- it would be a security problem. And DoCoMo is very sensitive to anything security-related."

Rescued by 3G -- Not Some pundits have pointed out that the new, hugely expensive 3G network could provide a pathway to expanded data use. No doubt, the spanking new W-CDMA system is the jewel in DoCoMo's crown. It's fast (220-plus Kbps and increasing), the handsets are gorgeous (onboard video cams, organic electroluminescent displays, USB port, et cetera), and there's a lot of promise for deploying enterprise applications.

Unfortunately, there are still major problems. Coverage is limited to the Tokyo area, calls are being dropped, and one early user wrote on a local Web site, referring to the NEC 3G handset: "On a full charge the battery doesn't even last one day -- and that was without making any voice calls; i-mode and i-Appli only." While the company is aware of all these problems, they will take time to fix.

Meanwhile, ARPU continues to drop while 3G contributes very little to increased packet usage. Also, packet fees on the 3G network will have to come down, but even increased usage won't make up for the reduced revenue. The major application now is email, which accounts for 46 percent of i-mode usage. Analyst Calder thinks DoCoMo would have to get a five- or six-fold increase in packet usage on 3G at the lower fee just to stay revenue neutral, but, he adds, that's unlikely to happen.

Just when the company couldn't have wished for any more bad news, yet another iceberg struck the hull. In August, the Japanese press reported that the world's largest wireless carrier, UK-based Vodafone, was in advanced negotiations to buy a substantial controlling stake in DoCoMo competitor J-Phone, of which it already controls some 46 percent. The move has to be a nightmare come true for DoCoMo, long used to the utter absence of any foreign competition in its domestic playground. The tie-up would pit a price-aggressive, marketing savvy, foreign-connected and newly revived J-Phone against DoCoMo. Analysts expect J-Sky subscribers to beat EZweb for the first time in August or September. "J-Phone has already won marketing share from KDDI," says the European investment bank analyst, adding that while there's no danger of DoCoMo's top slot being lost, the new J-Phone will present a significant challenge to Big D. So far, 2001 has been the year of discontent for i-mode and its operator parent. Network glitches, falling packet usage, spam, and the delay in the European rollout are all plaguing the giant carrier. Sadly, some, like Calder, see little chance for improvement, particularly in the falling data ARPU trend and its impact on rising marginal subscriber costs. "DoCoMo is unlikely to turn this around," he says gloomily. @

Note: The function "email this page" is currently not supported for this page.