ERIC GIRIAT

E L E C T R O N I C

C O M M E R C E

163

M-commerce:
An operators manual

Nick Barnett, Stephen Hodges, and Michael J. Wilshire

Mobile-telephone operators could compete on all levels of the mobile-commerce value chainbut they should think twice before they do.

pward of 300 million people more than own PCsacross the

world subscribe to mobile telephones. If, as forecast, this gure passes one billion by 2003, mobile telephones will be as common as television sets. Such a penetration rate combined with technology that provides for the fast transfer of data on mobile networks, standard protocols that deliver Internet-like services on smaller screens, and the very personal nature of mobile telephones will drive a mobile-commerce revolution (see sidebar, M-commerce: The next big thing, on the next spread). Within three years, the annual value of goods and services transacted over mobile networks could reach $13 billion, or 7 percent of all e-commerce transactions.1 Wireless operators are well placed to benet from this growth because they alone own and operate the mobile networks that make m-commerce possible. But as available bandwidth proliferates and the coverage capabilities of networks converge, the basic business of transporting data will become more competitive, and prot margins will fall. Hence operators are keen to seek value-creating opportunities beyond the simple transport of data.

Analysys (a European telecom consultancy) report, Mobile e-commerce, February 2000.

Nick Barnett and Stephen Hodges are consultants and Michael Wilshire is a principal in McKinseys London office. Copyright 2000 McKinsey & Company. All rights reserved. This article can be found on our Web site at www.mckinseyquarterly.com/electron/mcop00.asp.

164

T H E M c K I N S E Y Q U A R T E R LY 2 0 0 0 N U M B E R 3

M-commerce: The next big thing

Four main forces underpin the mobile-commerce revolution: third-generation technologies, the Wireless Application Protocol (WAP) and iMode platforms, handset penetration, and personalized services. The high speed at which Internet data can be downloaded is only one important characteristic of the new networks. In current wireless networks, most data communication, apart from the limited Short Message Service (SMS), requires a circuit-switched connection: a user must connect to a server to check e-mail, for example. This limitation has two drawbacks. First, users nd themselves on-line even when they are not sending data (while reading or composing e-mail, for example), so they pay higher costs and network capacity is wasted. In addition, since the connection can be initiated only from the mobile handset, asynchronous services, such as automatic forwarding of e-mail, are not possible. Like the wired Internet, GPRS networks use a connectionless (packet-switched) communications mechanism. Data are split into chunks called packets, to which an address uniquely identifying the destination is appended. This means that although a GPRS handset is, in effect, permanently connected to the network,

1. Third-generation technologies
Unfortunately, the current (second) generation of wireless networks and handsets supports data rates of only 9.6 kilobits per second, far below the 64-kilobit-per-second capacities of landline copper wires. That situation will improve this year, however, as GSM (Global System for Mobile Communication), the most common cellular standard, is extended by the General Packet Radio System. GPRS can support data rates of 112 kilobits per second, almost twice the rate of a standard computer modem and enough to support high-quality streamed audio. True third-generation networks, based on the UMTS (Universal Mobile Telephone System) standard, will raise the rate to 2 megabits per secondone-fth of the bandwidth available on the standard Ethernet in todays offices (Exhibit A).
EXHIBIT A

Third-generation technologies: A battle for bandwidth

Thousands of bits per second First generation: GSM1 2,000 High-quality video 1,000 Mediumquality video Graphics, audio Text Messaging SMS (Short Message Service) 2000 2002 2004 Second generation: GPRS2 Third generation: UMTS3

100 20 10 0 1998
1 2

Global System for Mobile Communication. General Packet Radio System. 3 Universal Mobile Telephone System.

M - C O M M E R C E : A N O P E R AT O R S M A N U A L

165

it uses network capacity only when packets are actually being sent.

telephones to access the Internet more often than they use personal computers.

2. The WAP and iMode

Two of the leading platforms for delivering Internet content to mobile telephones and other wireless devicesthe Wireless Application Protocol and iModewere designed to take into account the constraints of wireless communications: limited bandwidth and endsystem processing as well as a constrained user interface. Each platform denes a standard markup language that permits an applications user interface to be specied independently of the end device. The delivery of these services is independent of the underlying networking technology, so applications can be used on different networks, just as Internet applications can.

4. Personalization
The wireless Internet has three main features that permit mobile interactive services to be more personalized than traditional Internet applications are. First, mobile telephones are carried by their owners almost everywhere and kept switched on most of the time (especially in Europe, where mobile users arent charged for incoming calls). Consumers can thus not only gain access to wireless services wherever there is a network presence but also keep tabs on time-critical information, such as stock market reports or urgent messages. Second, wireless-network operatorsat least those using the GSM standardare uniquely able to determine the identity of a user. Since mobile telephones are not usually shared, and a personal-identication number often protects them, the telephone itself can be used as a means of identication. Finally, operators can detect a users exact location, enabling a whole range of new applications.

3. Handset penetration
The uptake of mobile telephones has been nothing short of phenomenal, and the trend is expected to continue (Exhibit B). Nokia predicts that within three years people will use mobile
EXHIBIT B

Europe picks up the phone

Mobile phone penetration in Europe,1 percent 100 90 80 70 60 50 40 30 20 10 0 1993
1

Compound annual growth rate = 36%

1994

1995

1996

1997

1998

1999

20052

Includes Austria, Belgium, Finland, France, Germany, Italy, Norway, Spain, Sweden, and the United Kingdom. 2 Forecast. Source: Financial Times ; Lehman Brothers; McKinsey analysis

166

T H E M c K I N S E Y Q U A R T E R LY 2 0 0 0 N U M B E R 3

In theory, mobile operators could compete at all levels of the m-commerce value chain, from the provision of basic technical services to the supply of lucrative, customer-facing content. The high stock market valuations of Internet-related companies, their powerful nancial muscle, and the absence of strong competition at Operators will have to make all levels in the new industry might difficult decisions about where even tempt them to try. The danger in the value chain to compete is that they will spread their skills and resources too thin. Moreover, given the rst-mover advantages associated with much Internet-related business, such companies risk forfeiting long-term shareholder value unless they concentrate on areas in which they naturally hold a strong competitive advantage.2 Operators thus need to make difficult decisions about which parts of the value chain to compete inand howand which parts to avoid.

The m-commerce value chain

There are seven links in the m-commerce value chain. At the bottom is transport: the maintenance and operation of the infrastructure3 that provides for data communication between mobile users and application providers. The second link consists of basic enabling services, such as server hosting, data backup, and systems integration. Vendors wishing to target wireless customers need these services to make products available via mobile telephones. Transaction support is the third link of the value chain. Many wireless services will require some form of paymentusually from the user to the service provider to pay for, say, books or CDs but possibly also in the other direction, for refunds or customer reward schemes. Transaction support provides the mechanisms for assisting those transactions, for security, and for billing users. The fourth link is presentation services. Providers convert the content of Internet-based applications, which are formatted in a standard known as HTML (HyperText Markup Language), into a standard such as WML
2

Many Internet-related businesses are characterized by low but rising entry barriers created through economies of scale in building a brand, network effects (the increasing utility of a service as more people come to use it), and economies of scale or scope in constructing the underlying technology platform. This infrastructure includes the backbone network, GPRS/UMTS (General Packet Radio System/Universal Mobile Telephone System) base stations, and the gateway between the mobile network and the Internet.

M - C O M M E R C E : A N O P E R AT O R S M A N U A L

167

(Wireless Markup Language), an HTML subset suitable for the small, lowresolution screens of wireless devices. Content that isnt already on the Internet can be formatted directly into a wireless standard. Personalization support is the fth link of the chain. One of the main value propositions of m-commerce is its ability to personalize applications for individual users. Providers that wish to offer the best m-commerce services need information such as the users name, address, location, and billing details (the number of a credit card or a bank account, for example) and evenbecause the size of the screen affects the kind of information that can be viewedthe type of device used to connect to the service. Companies that can provide such information will form a valuable link. User applications are the sixth link of the value chain. Possible applications range from those currently available on the wired Internet (including banking, book purchasing, e-mail, news, and travel) to new services designed specically for mobile consumers (information about where to nd the nearest coffee shop, for example, or the Portals are particularly important automatic notication of nearby for m-commerce because mobile friends). At the highest point in the telephones have small screens chain are the content aggregators: businesses that design and operate portals, which provide information in a category or search facilities to help users nd their way around the Internet. This function is particularly important for m-commerce because mobile telephones have small screens and limited input mechanismsnotably, no mouse and a non-QWERTY keypad. Users will want portals that simplify the search, avoid throwing up too much information, and require minimum input.

Opportunities for operators

Few industries offer the opportunity to compete in every link of a value chain, but mobile operators have the necessary technical skills, particularly at the lower end. They also have the billing systems for payments, as well as brands that potentially position them well for customer-related activities higher up the chain. In addition, their control of the wireless infrastructure and their ability to congure subscribers handsets to make themselves the default Internet access provider give them the power to limit subscribers access to competitors services at least initiallyand therefore to build their own branded ones. Early indications are that operators will indeed try to use these advantages to capture value in many parts of the chain. Sonera and Cellnet, for example,

168

T H E M c K I N S E Y Q U A R T E R LY 2 0 0 0 N U M B E R 3

have already launched portals; Vodafone AirT ouch and Cellnet have application services; Telenor and Sonera offer transaction facilities; and NTT DoCoMo supplies basic enabling services, presentation services, and personalization facilities. A glance at recent stock market valuations of the Internet counterparts of the companies supplying these services shows the potential attraction of such businesses: Yahoo! (content aggregator) is valued at $89 billion, Amazon (applications) at $21 billion, Exodus Communications (enabling services) at $25 billion, and Engage Technologies (personalization) at $5 billion.4 But a glance at the Internet also reveals it to be a specialists world where erce competition forces companies to excel. The Internet players will compete with mobile operators, exploiting the distinct advantagessuch as brands and experience the Internet players have already developed on the World Wide Web. T o withstand the competition, mobile operators should concentrate on the four areas in which they have a strong advantagetransport, personalization support, content aggregation, and transaction supportand move quickly to capture these opportunities by adopting four strategies.

1. Drive traffic growth by offering a wide variety of data services

Data traffic (mostly driven by the simple text-messaging service called the Short Message Service, or SMS) already contributes 10 percent of the revenues of some wireless operators and, according to forecasts, will soon overtake voice traffic as the main source of revenue for mobile operators as a whole. But as volumes rise, prot margins fall: in the United Kingdom, call charges have dropped by about 77 percent in ten years.5 In response, some operators have used their control of the network infrastructure to try to lock in value at stake elsewhere in the chain. By presetting their subscribers telephones to make themselves the default Internet access provider and blocking unauthorized services, operators have the opportunity both to charge application providers for access to their subscriber base and to build their own branded services. This walled-garden strategy might provide higher revenues in the short term, but in the longer term it is awed. First, it fails to maximize demand for data traffic, because users may be barred from their favorite on-line services. Second, it runs the risk of driving away those dissatised subscribers, who may switch to a competitor to get exactly what they want.
4 5

As of March 14, 2000. These Analysys gures are based on the average monthly bill for a high-usage business customer taking advantage of the cheapest available tariff.

M - C O M M E R C E : A N O P E R AT O R S M A N U A L

169

Since consumers increasingly choose their wireless providers on the basis of the data services available, this second point is an important one. McKinsey research in the Asia-Pacic region, for example, indicates that half of all subscribers (and up to 70 percent of high-value ones in some countries) would switch operators to get better wireless data services. It follows that the rst operator in a region to offer unrestricted access to data services could capture a lucrative portion of the subscribers of other networks. The best way to go on proting from transport is therefore to increase traffic on the network. This can be done only by offering subscribers access to the widest possible range of data services, an approach that has been validated in the Japanese market by the success of NTT DoCoMos iMode service. The way to prof t from transport is Launched in February 1999, within to increase traffic on the network a year it was offering more than by offering subscribers access to 3,000 Internet-like services and had the widest range of data services attracted three million fee-paying subscribers (rising by 10,000 a day on average). Average revenue per subscriber also rose by about $120 a year$30 from at-rate subscription fees and $90 from increased traffic revenue.

2. Attract content providers by offering good user information

Customer data are a valuable asset in the off- and on-line worlds alike, for the more a vendor knows about a customer, the more it can personalize the information or service it provides. The end result is likely to be a more valuable service, reckoned both by price charged and by the vendors ability to attract and retain customers. In the wired world, Internet vendors use various techniques to personalize a customers visit to their sites. These techniques range from cookies pushed to the users computer to ensure that repeat visitors are recognized and appropriate data recalled, on the one hand, to complex collaborative lters that predict the preferences of customers from their behavior, on the other. (An example is Amazon.coms customers who bought this book also bought . . . service, which exploits the purchasing patterns of similar customers to suggest books or CDs that users might enjoy.) In the wireless world, personalization will be even more important, for screens are small, limiting the amount of information that can be shown, and information can be tailored with the help of a few personal details. (A search for a restaurant, say, might be rened by the users location and

170

T H E M c K I N S E Y Q U A R T E R LY 2 0 0 0 N U M B E R 3

spending bracket.) The kind of device that is used will also determine the kind of information that can be sent: a stock quotation service, such as Yahoo! Finance, could supply shareprice graphs to devices with sufficiently large screens. Application providers will probably

be attracted to those networks whose operators can supply the most useful data on customers

Input capabilities too are limited, which means that users will benet if input forms can be lled in automatically on their behalf. A CD vendor, for example, could simply ask customers to verify payment information and a shipping address rather than have them ll out forms from scratch. Mobile operators hold plenty of this kind of personal information on subscribers and are well placed to supply application providers with the data they require (subject to legal restrictions, which differ from country to country). Mobile operators can also use this information to drive traffic growth, for application providers are likely to be attracted to the networks whose operators supply the most useful customer data. A larger number of application providers will in turn attract more customers to the network. The upturn in traffic will attract more application providers, more customers, and so on in a virtuous circle.

3. Assume the role of a wireless portal

Like portals in the wired world, wireless portals have a degree of control over what users see on the Internet, so the portal provider can charge service providers and advertisers high fees. Given the projected penetration rates of mobile devices and mobile consumers increased reliance on portal services, many observers expect wireless portals to be as highly valued on the stock market as their established Internet equivalents.6 Operators enjoy some natural advantages in providing portal services. First, the operators can control the conguration of telephones for all their new subscribers and thus make their portals the default start screen.7 Second, the operators information about subscribers permits them to tailor the content provided. Indeed, the way portals use this personal information could turn out to be the main factor distinguishing good from mediocre portals. (A good one, for example, wouldnt require users to ll in personal details for each site they visited an off-putting task on a small keypad.) But
6

Goldman Sachs recently valued Vodafones Multi-Access Portal at almost $29 billion, for example a valuation based on Goldmans predictions of the potential revenue sources for a wireless portal. Although the user can modify this, doing so requires some effort.

M - C O M M E R C E : A N O P E R AT O R S M A N U A L

171

Internet experience indicates that operators will have to move quickly to succeed as portal providers. In the United Kingdom, Freeserve stole a march on competitive portals by offering free Internet access and retains a strong first-mover advantage. Operators must also decide whether to build the portal alone or with a partner. Despite some natural strengths as portal providers, operators have so far shown limited skill in selecting, aggregating, and customizing content, for though they are good on the technical side, they are light on the kind of media and marketing skills that characterize successful Internet portal providers. And the value at stake at this level of the chain ensures that operators will face stiff competition, not least from existing Internet portals with strong brands, a large subscriber base, and plenty of experience. Several established companies, including Microsoft, Yahoo!, Excite@Home, and America Online, have already launched wireless portals. Operators have three broad options when considering a portal strategy. They can operate their own portal under their existing brandthe approach taken by operators such as Telia Mobile and NTT DoCoMo, which identied the potential of m-commerce long before others did. They can outsource the provision of the portal to one of the new breed of wireless portals, such as AirFlash or Room33, perhaps rebranding the service under their own names. Or they can form a co-branding partnership with an established portal. The latter two options are the safest bet for less advanced operators, particularly in countries where existing Internet portals will offer strong competition. The benets to both parties are clear: the operator gains access to the aggregation skills of the portal; the portal gets hold of the operators customers and information about them.

4. Provide transaction support with a wide range of payment choices

All operators have systems to bill and charge their subscribers. The same systems can be used to bill subscribers for goods and services sold by third-party vendors on the network and to levy a per-transaction charge to vendors. Some operators have already spotted this opportunity. Sonera makes it possible for subscribers to buy soft drinks from vending machines by dialing a number, and the cost is then charged to the subscribers mobiletelephone account. Telenor is experimenting with a similar system that permits customers to book cinema tickets, while subscribers to NTT

172

T H E M c K I N S E Y Q U A R T E R LY 2 0 0 0 N U M B E R 3

DoCoMo can add all of their m-commerce transactions to their monthly mobile bills. Not surprisingly, this role is being contested by credit card companies that have their own billing systems, established relationships with service providers, and expertise in making customers credit and transactions secure. These credit card companies are To offer customers a choice of payment methods, smart operators even threatening to bypass operators will probably join with one or more altogether by forming relationships directly with handset vendors. Visa banks or credit card companies and Nokia, for example, are jointly developing a system permitting consumers to pay for goods and services using bank information stored in a telephones memory. Yet mobile consumers will probably demand a choice of payment mechanisms: their credit card account, their bank account, and their monthly telephone bill, for example. To offer customers a choice of payment methods, smart operators are thus likely to join up with one or more banks or credit card companies. Sonera recently did precisely this with MasterCard.

Where not to play

In the areas of basic enabling services, presentation services, and user applications, wireless operators hold no natural advantages and are likely to face strong competition. Because of the similar skills required in the wireless and wired worlds, existing Web-hosting companies and system integrators that already focus narrowly on these businesses will probably dominate basic enabling services. Presentation services will typically be provided by software companies that convert standard HTML into a form suitable for wireless devices and by Internet design rms that offer direct conversion services. Existing Internet content providers will in any case have a strong lead in user applications, since the providers quickly make existing services available to mobile consumers. The most successful applications, moreover, tend to be developed by focused entrepreneurs who dedicate themselves to the task. One question remains: how broadly should mobile operators compete in the four areas in which they have a competitive advantage? Given the likely competition, operators should consider placing their big bets in no more than two areas, depending on where they have the strongest advantages as

M - C O M M E R C E : A N O P E R AT O R S M A N U A L

173

measured by knowledge, skills, or the absence of competition. Elsewhere, they should seek to dilute their level of risk by forming strategic partnerships or joint ventures. It is important to remember that m-commerce is developing fast and that mobile operators need to establish an early lead to capture maximum longterm value. Focus is one of the surest ways to build on what is already their strong starting position in this market.