47275 Going Cashless at the Point of Sale: Hits and Misses in Developed Countries S toring value electronically ... sending value cards have failed, because customers often are electronically ... many people living in developed not convinced of the need or practicality of these countries take these things for granted because systems. Europe tested the market for these services making electronic transactions are part of everyday early on--and collected some high-profile failures routines for them. After all, who would think twice in the process, from cash-substituting smartcards NOTE about making a payment or getting cash from a in the second half of the 1990s (Mondex, Proton) debit or credit card? to interoperable mobile payment platforms in the early 2000s (Simpay, Mobipay). Debitcardsareindeedbecomingastandardpayment instrument for people with a savings account. Debit The more developed markets in Asia--Japan, cards have achieved critical mass adoption in most Hong Kong, Korea, Singapore, and Taiwan--have developed countries. Their further spread is mostly taken the lead in devising new schemes and are, limited at this point by three factors: in fact, meeting with some success. So against overwhelming failures in Europe, we can point to · The penetration of bank accounts among the underwhelming successes in Asia. population. · The weak business case for deploying a Although these developed country sufficiently dense network of acceptance experiences may not directly translate terminals (automated teller machines [ATMs] or into lessons that can be used in developing point-of-sale [POS] devices) in environments with countries, this paper better informs us low economic activity or population density. about what may or may not be possible · The cost of communications underpinning the and may or may not be different in the real-time authorization of payments, in markets developing world context. with limited communications infrastructure or for very low-value transactions (for which the There may be factors specific to Asia at play in communications cost as a percentage of the explaining the more positive experience in Asia-- transaction cost may be too high). customers' fascination with new technology, FOCUS the importance of conveying innovation in the These are important limitations in many developing branding strategy of mobile operators, the low countries, where the spread of banking services and penetration of credit cards. Some of the Asian infrastructure is often restricted by socioeconomic formulas are now finding their way back to Europe. stratum and geography. Even in developed Hong Kong's success with transit cards (Octopus) No. 51 countries, these limitations may create niche is also proving its mettle in London (Oyster). December 2008 opportunities for alternative electronic payment Japanese and Korean operators (NTT DoCoMo (e-payment) schemes to exploit the gap between the [DCM] and SK Telecom, respectively) are slowly Ignacio Mas and informality of cash and the heavier communications proving the benefits of mobile phones with very Sarah Rotman infrastructure required for debit cards. short-range communications technologies1 for use with in-person mobile payments, and many Yet, many initiatives that have sought to push European operators are now anxiously awaiting the frontier of electronic money (e-money) and the spread in Europe of phones with embedded payment devices to drive out cash beyond debit NFC capabilities. 1 We use the term "very short-range communications" to include infrared, radio frequency identification (RFID) or near-field communications (NFC) technologies. 2 This paper reviews some of the bigger failures reliability of the new system. Mondex debuted and some of the more promising experiences with much fanfare in several countries, and then in the use of smartcards and mobile phones as languished. payment platforms in developed countries. We selected just a few examples--from dozens of Octopus is also a stored-value smartcard, but it possibilities2--and did not delve into much detail was launched with a much narrower ambition: as a on any given scheme. Beyond telling the stories public transport ticketing system. It excelled at this of these ventures, our objective is to extract some niche application, in no small part helped by very lessons behind the failures and the successes. favorable ticket pricing schemes. Now Octopus Although these developed country experiences is growing out of the subways and ferry stations may not directly translate into lessons that can be where it developed and is becoming an accepted used in developing countries, this paper better form of payment at many retail establishments. informs us about what may or may not be possible This model is being replicated by mass transit and may or may not be different in the developing consortia in many countries. world context. Mobile operators facilitating A Look at Three Broad existing payment instruments Approaches Mobipay was an attempt by mobile operators in We discuss three broad approaches, and in each Spain to establish the mobile phone as a virtual case we look at two providers who met different wallet--the single interface through which degrees of acceptance in the marketplace. people could access any of their usual payment Although our primary interest is with payment mechanisms. By integrating the wallet into the through mobile phones, we start with two cases phone, the latter would emerge as an ever more that use smartcards, because these share many of personal and personalized expression of the the same characteristics and issues as payments mobile operators' service. Unlike Mondex and through mobile phones. Octopus, Mobipay did not require customers to accept a new form of payment (although it did add Smartcard-based electronic- the option of operator billing). It simply sought to cash providers change how users accessed their existing payment options. But even this proved difficult: Mobipay Mondex was a stored-value smartcard that allowed failed to spur sufficient interest, and only a small users to make low-value, in-person payments. share of customers saw enough benefit to actually It could be loaded from regular bank accounts try the system (beyond using it as a mechanism for at specially equipped terminals, and value was electronic airtime top ups). stored on the card itself without back-up in a server, which made much faster, offline payments SK Telecom in Korea followed a similar approach, possible. Taking on cash in its home front--at but simplified the procedure for customers: it the shop--proved too much. Customers failed to focused on enabling credit card payments only. It appreciate strong advantages over paying in cash used contactless capabilities so that phones could and remained concerned about the security and communicate directly with merchant terminals 2 In Europe, the other more substantial e-wallet schemes are Belgium's Proton, the Dutch Chipknip, Austria's Quick, Germany's GeldKarte, and Sweden's CashCard. In Asia, other noteworthy initiatives include CashCard in Singapore, Touch `n Go in Malaysia, and Chugwa Telecom's emome in Taiwan. 3 simply by waving them over the terminal, thereby In Japan, the leading mobile operator, DCM, also reducing the need for cumbersome data entry by sought to create an operator-centric transactional users at POS. It also permitted offline transactions and payments ecosystem, but it exercised a much by storing credit card details directly in a special tighter degree of control over the platform, focusing chip on the mobile phone. This functionality not on interoperability with other operators but on came at a price: it took a long time before it was domestic payments. It also sought to increase the interoperable with other POS terminals, and it customer's payment options. The service, Osaifu- required special phones, which slowed take-up. Keitai, allows customers to charge goods and But more fundamentally, the value to customers of services to their mobile bill, against DCM-backed not carrying a card in addition to a mobile phone e-money, or to DCM- (and third-party) issued credit remains unproven. cards. DCM understood the importance of driving acceptance of the service with both customers Mobile operator-centric and merchants, and it developed a sophisticated payment schemes ecosystem-building strategy. Nevertheless, customer take-up and especially use remains The leading mobile operators in Europe sought relatively low. to go beyond channeling customers' payment instructions. They devised a scheme called Simpay Back to debit and credit cards to put their own billing platforms at the heart of a new small-value payments system.3 Goods Is the track record of introducing new retail purchased through a mobile phone or a personal e-payment instruments and devices in developed computer (PC) would be billed to the customer's countries really so bland, if not bleak? Not at all. mobile operator account (prepay or postpay). By Consider the inexorable rise of credit cards in the bypassing traditional bank and credit card payment United States in the 1980s and the catch-up with instruments, participating operators sought to debit cards in Europe in the 1990s. These payment lower the cost of such small payments, while at cards have largely replaced checks. They are fast the same time strengthening their direct customer becoming the real competitors with cash today in billing relationships. physical (in-store) payments, and they dominate in remote payments (using the phone or Internet). Learning from the Mondex experience, they initially targeted online payments where cash is Bolt and Humphrey (2007) report that store-value not an option. Its use as a cash substitute would cards accounted for just over 1 billion euros in then emerge naturally over time as customer payments across 11 European countries in 2004 familiarity with Simpay grew. Simpay never had the (by which time their novelty had largely gone). Yet opportunity to test its proposition with customers. in that year the value of debit card transactions was It failed to navigate through the increasingly estimated to be over 1 trillion euros. Debit cards diverging strategic interests of its own backers, have indeed morphed into what Mondex hoped and the multioperator, pan-European nature of the to be, and customers now use them for even very service proved more technically complex than the small transactions (less than 5 euros). market opportunity justified. 3 Although there is no strict definition for what constitutes small-value payments, they are generally held to be below 10 euros (US$14). Jones (2008) reports that as many as one-third of transactions in Europe are in fact under 5 euros. 4 Figure 1: Categorizing the cases considered in this paper Mobile phone-based Smartcard- Led by single Led by industry based company consortium Paying from existing Debit Card Moneta Mobipay accounts New Mondex Osaifu- payment Simpay Keitai vehicles Octopus Yet the topic of alternative e-payments devices and debit cards at a much lower unit transaction cost. instruments will keep coming back. We are far from Or perhaps banks will continue to fill the void, a payments end-game, for two main reasons: finding ways to lower the cost of card transaction processing and launching their own alternative · Debit, and especially credit, cards are still prepaid products. Or, to paraphrase Jones (2008), costly instruments. Detailed costing studies of we will continue to lose the war on cash. Whatever alternative payments instruments in Belgium the eventual outcome, we will continue to see no and the Netherlands reveal that cash is still more shortage of new technologies and new business economical than debit cards for purchases below models being thrown at the "cash problem." 10­12 euros4; the cut-off would be higher still for credit cards. Much of the technical literature Summary: A classification of the laments that people do not have adequate payments systems considered incentives to use the most economically appropriate payment mechanism for each type Our six cases (seven, if we include debit cards) are of transaction, which perpetuates the use of very diverse in their scope, approach, intended cash and pushes more costly alternatives, such as applications, and nature of backers. Figure 1 credit cards, instead of promoting the emergence offers a framework to visually represent the major of new very low-cost payment options. differences along three key dimensions: whether · E-payments are now firmly based on bank they work from existing bank accounts or new systems--credit cards, debit cards, Internet forms of payment (e-money, operator billing), the banking. There is a need for electronic payment instrument used (card or mobile phone), instruments that target the unbanked. and whether they are led by individual companies or based on interoperable industry consortia. So there may be room for a nonbank-based The position of each scheme within the matrix payment instrument that has all the advantages of represents the primary application of the scheme. 4 These calculations take into account (i) social costs (i.e., net costs to all parties involved, including customer, bank, and merchant) of the various instruments and (ii) variables costs (i.e., not including the cost of payment terminals and nonscalable back-office infrastructure). See Brits and Winder (2005) and National Bank of Belgium (2006). 5 Debit cards are just a method for accessing liquidity The concept was invented in 1990 in the United in your normal bank account. Mondex is a new Kingdom. Mondex UK was officially incorporated form of liquidity (e-money) intermediated (i.e., sold in 1992 by two U.K. banks (NatWest and Midland), to users) by banks and linked to bank accounts. and the umbrella company, Mondex International, Octopus is a self-standing reloadable store-of- was established in 1996. MasterCard acquired a value system operated by a single company; it may 51 percent ownership of this company later in the be topped up through a variety of mechanisms, same year. including through cash machines. How it works On the mobile phone side, Moneta is primarily a mechanism for mobile-enabling credit card Under the Mondex system, customers are issued payments (Moneta Chip), although it originally a smartcard, which is a plastic card with an also included an e-money option. At the other integrated circuit or chip from which data can be extreme, Simpay is a system designed primarily read and updated by card-reading devices. The to support operator billing (charging to prepay or chip stores the money value, but this is not linked postpay mobile accounts) for low-value payments. to any bank account, and the information is not It constitutes an alternative settlement platform backed up on a server. Thus, the money value is for customer transactions. Mobipay is a mobile irretrievably lost if the card itself is lost, and the wallet service enabling a range of bank payment card provides a much higher level of anonymity instruments (debit, credit), as well as operator billing for users. to a limited extent. And Osaifu-Keitai is the broadest in terms of the range of instruments it supports: Consumers load Mondex value by transferring credit cards, e-money, and operator billing. money from their (separate) bank account using an ATM or a Mondex-enabled telephone, which A cash substitute: Mondex has a card reader and is connected to the Mondex in the United Kingdom system. The float is kept by the Mondex issuer (a private company set up by Mondex International What is Mondex in each country), from whom participating banks would "buy" Mondex value to meet their Mondex is electronic cash on a card. It was customers' requirements for Mondex value. designed to mimic cash, particularly small notes and coins (e.g., micropayments), so transactions Value can be transferred between two cards by with Mondex had to be extremely fast and had to inserting them into the same card-reading device. incur no transactions cost. Accordingly, Mondex Most commonly this would be a POS terminal at was conceived as an offline transactional method a store (for transactions between consumers and that does not require clearing systems to support retailers). It could also be a pocket-sized, dual individual transactions. Mondex-type systems are card-slot consumer device (called a wallet) that generally called "e-purses." Many such systems enables person-to-person (P2P) transactions. While emerged across Europe in the late 1990s: the Proton Mondex was conceived primarily for in-person system backed by Visa in Belgium and the Danmont transactions, with the two cards coming together system in Denmark were other early projects.5 in a single device, Mondex value could also be 5 This section is based on Clarke (2006), Stalder (1998), and Stalder (2002). 6 transferred between remote wallets through Customers will adopt a new payment service or Mondex telephones. Subsequently, Mondex technology when (i) it provides clear benefits developed a card-reading device that could be relative to current alternatives and (ii) they can attached to PCs to enable Mondex value transfers trust it based on a clear understanding of the risks via the Internet. involved. Mondex failed to convince the public on both grounds. Because the monetary value is stored directly in the card, the card-reading device does not need Utility any communications capability: transactions Customers are culturally accustomed to using between cards are final and are not authorized or cash, and hence a cash-substitute product does validated by a third party. There is no password not need to be simply "as good as cash," it needs authentication of the card owner. Thus transactions to be better to justify the switch. The utility of the can take place very fast, and no communication new payment mechanism can be evaluated on two cost is incurred. And because the system does core fronts: convenience and ubiquity. not operate from bank accounts, transactions between wallets do not create a need for clearing · Convenience. Users did not find using Mondex and settlement systems between banks, and hence faster than using cash. The consensus now is that there is no back-office cost. Because the value Mondex went head-to-head against cash in the is stored in individual cards, there is no central hardest market--small in-store transactions-- accounting system. Mondex's security architecture without having a clear advantage. It did not ensures that value is added to one card at the offer a very practical solution for the P2P same time that value is reduced from the other payments market, because it required customers card. The value is stored in hardware (the card's to buy a specialized wallet device. Mondex chip) rather than in software on a server, which was subsequently refocused toward online offers a higher level of intrinsic security. (e-commerce) rather than in-store payments, but in the end it was overtaken by simpler schemes, The market verdict and lessons learned such as PayPal. · Ubiquity. The usefulness of the card depends Extensive field tests were conducted in Swindon, on the adoption of card-reading devices by in the United Kingdom, during 1995­97 and in merchants. The customer proposition for Guelph, Canada, in 1997­98. Both involved a very merchants is in principle much more tangible strong marketing push to get a critical mass of than for customers, insofar as it reduces merchants to take up the card-reading devices. In cash handling costs and permits automated the Swindon trial, 14,000 cards had been issued accounting of transactions. However, the store by the time the trials were discontinued after 3 will realize these benefits only if a significant years, compared with the uptake of 25,000 cards proportion of customers adopt the new payment that had been anticipated for the first year alone mechanism. Mondex was never able to create (Van Hove 2005). Use of the cards turned out to enough critical mass of card holders wanting be much more disappointing. Although this service to pay for goods at stores using their Mondex still lingers in a few countries, it never met much card, with the result that merchants did not push market success. for this form of payment--even if it was in their collective longer term interest. 7 Risks of such data by merchants (who own the card- Through the trials, customers remained unsure reading devices) and by Mondex itself. Mondex about several aspects of the service. Using Mondex was unwilling to disclose exactly what data were seemed to incorporate substantial risks that people being captured, to protect the confidentiality did not have to worry about when using cash. of its security system. Whether justified or not, Mondex was never fully able to shake the · Loss of value. Although both cash and Mondex public's perception that information was being value stand to be lost irretrievably, Mondex is captured. electronic and hence more vulnerable to physical · Sustainability of Mondex value. Just what were damage (notes and coins can survive being users who loaded their Mondex card acquiring? run over by a car, for instance) or malfunction. Because it is not bank-based, Mondex value, Because the value is not backed up on servers, which is issued by Mondex itself, was not Mondex could not guarantee restitution of value guaranteed by participating banks nor did in the case of a faulty chip, which left customers it fall under government deposit insurance wondering about the reliability of the system. schemes. What protection would users get Mondex was less safe than cash, and even against a potential collapse of Mondex? The less so than other electronic forms of payment answers to these questions were never fully whereby if you lose your card you don't also lose articulated, and customers were kept guessing the balance in your account. on the implications to them of a failure of the · Security. Because Mondex transactions happen system. The public distrusted the new form of offline, are not recorded in central accounting money, which was only reinforced by the security repositories, and do not leave (much of) an concerns. audit trail, it is very hard to detect cards that have been successfully tampered with. Thus, the A transit card: Octopus security of the system needs to rely on the chips in Hong Kong on the cards being tamper-resistant, without there being much of a possibility of using "fall What is Octopus back" intrusion detection methods. However, no hardware can be made absolutely tamperproof, In 1979, Hong Kong's Mass Transit Railway (MTR) so security breaches, though unlikely in launched a prepaid card with a magnetic strip as practice, could have very serious repercussions a ticketing system for use with its rail services. In if they occurred. Thus, despite the high security 1994, Creative Star (renamed Octopus Cards Ltd. standards embodied in the Mondex system, it in 2002) was formed as a joint venture between continued to be dogged by questions about MTR and four other public transport operators whether the security could be ensured longer in Hong Kong to make it an intermodal ticketing term. system (i.e., including buses, ferries, subway, · Privacy. Although it was initially marketed as etc.). The Octopus card, a contactless smartcard a completely anonymous transactional system, based on Sony's FeliCa chip, was introduced in some transaction details are captured by the 1997, replacing the old magnetic strip cards. The card-reading devices. There were concerns by card does not need to be physically inserted into public privacy advocacy groups about the use a device to be read, which makes payment very 8 convenient for users in a hurry: all they have to that the older Queen's Head coins in Hong Kong do to pay the exact fare is to swing their purse or would appreciate, causing people to hoard the handbag near the card reader. coins. The value is stored securely in the card itself. The Although the Octopus card was originally designed card can be personalized for an extra charge with a as a single-purpose card, Octopus moved to retail photo, and personal data can be kept on record. If payments and launched the Octopus Retail card. a personalized card is lost or stolen, the customer In 2000, Octopus obtained a special purpose can reclaim the remaining value of the card, and deposit-taking company authorization from the the original card will be blacklisted to prevent its Hong Kong Monetary Authority to allow it to use. develop different applications. The use of the Octopus card was subsequently extended first Value can be reloaded at transport stations, over to transport-related areas (e.g., parking, shops the counter, or by feeding banknotes into value- at transport terminals) and then, from 2000, to adding machines. More recently, cards also can vending machines and unrelated retail outlets. be reloaded online or at 7-Eleven stores and In November 2004, all parking meters in Hong electronically at ATMs. There is also a convenient Kong were converted to work on the Octopus automatic top-up option directly from an account card system. Interestingly, taxis do not accept (the card is automatically reloaded against a credit Octopus cards, because Octopus does not have card account with HK$250 or HK$500 as soon as an automatic account updating system, so drivers the card value falls below zero). There is a maximum would have to return to the office every day for value storable on the card (around US$130). The accounting purposes. Merchants pay a fee to float is held by Octopus Card Limited, and the full Octopus for each transaction customers make with value is deposited in banks. their Octopus card. At the time of launch, acquiring an Octopus The Octopus card has also been used for card required a deposit of HK$50, which created multipurpose ticketing (e.g., combining station car widespread resentment with the new payment parking, the train ticket, and the ticket to an event mechanism. However, adoption was driven by (i) into a single fare). The Octopus Rewards scheme, very rapid conversion of all turnstiles to the new added in November 2005, provided benefits at system; (ii) a short phase-out period for the old participating retailers. And in July 2008, Citibank ticketing system of only 2­3 months; and (iii) a launched a combined contactless Octopus travel pricing scheme of the only remaining ticketing and credit card. alternative--a single trip ticket--at a much higher price. This amounted to a compulsory conversion The Octopus card is evolving into a multiapplication by all transport users, such that within 3 months, platform. Personalized cards can be used as a key three million cards--a number equal to half the card for access to residential and office buildings residents of Hong Kong--were sold (Siu 2002). or in a university environment to record student Adoption of the card was also helped by a coin attendance and manage a student's library loan shortage in 1997, at the time of the transfer of account. Octopus has also extended the range of Hong Kong to China, that resulted from the belief physical formats in which it is embodied: Octopus 9 sells watches and mobile phone covers that gasoline stations by simply waving the fob key over function as an Octopus card. the payment terminal. In the wake of Octopus Convinced of its own success, Octopus created a wholly-owned subsidiary called Octopus Transit operators around the world are following Knowledge Limited, to provide international in Octopus's tracks. Transactions that could be automatic fare collection consultancy services. conducted by the Oyster card, issued by Transport for London, was extended to include low-value The market verdict and lessons learned purchases at retail stores in January 2006. Transit cards have the potential to go cardless: The Octopus system is now a widely used Oyster payments proved to be the most popular electronic cash system. By mid-2008, there were application in trials conducted by Nokia and U.K. over 17 million Octopus cards in circulation (which mobile operator O2 in early 2008 using new mobile is more than twice the population of Hong Kong), handsets enabled with NFC, with 89 percent of with more than 10 million transactions, worth trial participants saying they would like to keep the HK$85 million, processed daily (Citi 2008). The functionality on their handsets after the trial (Week cards are used by 95 percent of the population in Wireless 2008). of Hong Kong aged 16 to 65; the average user stores around HK$63­65 on the card. It is evolving Many cities are now adopting smartcards for from a transit card to a retail or "city card." By mass transport ticketing, such as SmarTrip in late 2006, the card was accepted by 3800 shops Washington, D.C.; T-Money in Seoul, Korea; operated by 379 companies. According to De Jong Snapper in Wellington, New Zealand; Suica in (2006), only 76 percent of transactions were for several metropolitan areas in Japan; and Easycard transport services. "In the company's lifetime, its in Taipei. focus has shifted from delivering a not-for-profit, cost-reducing technology to being a marketing In a similar vein, many e-money cards have emerged organization today" (De Jong 2006). out of road-pricing schemes. Singapore's CashCard was the mandated instrument for road pricing in Papadopoulos (2007) presents a much more the island and then expanded to other low-value skeptical view: the "head start" from its forceful applications.6 In the United States (EZPass) and introduction as a transit payment card "did not in Australia (e-TAG), drivers rely on smartcards materialize in higher use or quicker development to zip through tollbooths, but that success hasn't for Octopus Retail than other e-money projects traveled far beyond the highway. elsewhere." This "provides a note of caution [;...] it may be easy to expand as a single payment medium Smartcards with a short-range connection have in a protected environment, where the issuer is also been tried in unattended POS environments also creating and controlling the demand of its beyond mass transport systems. Exxon in the product and is a completely different challenge to United States introduced this capability as a simple compete in unequal terms against the established key fob (a small device that can be attached to use of cash in retail." a key chain) with which customers could pay at 6 Based on the success of the CashCard, the Government of Singapore developed a plan called Singapore Electronic Legal Tender to grant legal tender status to some form of e-money and to drive out the use of cash entirely. There is not much information on the status of the initiative, which seems to have been relegated in importance and time. 10 Van Hove (2005) notes that Mondex failed to take Interoperable mobile wallet root in Hong Kong, whereas Octopus has taken supporting multiple payment root. Octopus had four distinct advantages over instruments: Mobipay in Spain Mondex: What is Mobipay · Having signed up all the main public transport companies in the territory, Octopus had a de- Mobipay is a mobile payment mechanism that allows facto monopoly with a large user base--mass customers to pay for goods through their mobile transit users. It was able to attain substantial phone using a range of payment instruments: credit customer adoption in a remarkably short time cards, debit cards, or the operator's account. It by installing the new card readers at all locations supports both in-person and remote payments. under its direct control and using relative pricing The platform is open to any mobile operator or of the various access methods (Octopus vs. payment instrument issuer in Spain. single-journey tickets). In effect, users were left with no alternative. Once the Octopus card had The system is based on a cooperative model attained mass use, it was then able to extend its between mobile operators and financial institutions. use outside of its captive market in normal retail Mobipay España S.A., which operates the system, outlets. In contrast, Mondex had to prove that it was formed in July 2001. It is co-owned by the was better than other payment alternatives on three major mobile operators in Spain (jointly the basis of its own merits and did not have the owning 40 percent), as well as by over 30 financial option of manipulating pricing at retail stores to institutions (jointly owning 48 percent), and the provide incentives for its use relative to other three national card payment processing companies means of payment. (jointly owning 12 percent). At its inception, system · Octopus focused on replacing cash in unattended development received financial support from the POS--ticketing machines. The ability to present Spanish Ministry of Industry. Mobipay España exact fares at all times through a smartcard at S.A. was actually formed as the merger of two unattended machines offered a big convenience competing projects: Movilpago (backed by a major factor for users. In contrast, at least initially, bank, BBVA, and the incumbent mobile operator Mondex attempted to replace cash in stores, Telefónica Móviles) and Pagomóvil (backed by where not having the exact change is much less another leading bank, BSCH, and the two other of a bother for customers. mobile networks). · Octopus's contactless features made it extremely convenient for users--indeed, faster than using A separate company, Mobipay Internacional S.A., cash. The cards do not have to be inserted was created to promote this payment solution in and, in fact, generally do not even have to be foreign markets, especially Latin America. However, withdrawn from wallets and handbags, to be unlike the Spanish entity, the international one read by the card readers. is controlled by just one telecommunications · The card's personalization feature means that company (telecom)--Telefonica Moviles--and value can be retrieved by users who lose their one bank (BBVA), which should give it a stronger card or have a faulty card. strategic coherence. It has launched in Mexico through FINPE, an organization created in 2004 11 by a consortium of banks and credit card operators There are three main ways of initiating a payment. with the sole purpose of promoting the use of For smaller, in-person transactions, customers can e-payments in Mexico. give their phone number to the merchant, who will then issue the payment request. Larger retailers How it works can use a special barcode reader that can acquire the customer's mobile phone number directly by Each Mobipay customer gets a virtual wallet, reading a tag on the customer's phone, which which can contain up to nine different payment reduces the possibility of error. For purchases from instruments. Each time the customer wants to machines or for remote purchases, the customer make a payment, the system will ask which of the can enter a transaction code that identifies the available payment instruments the user wishes product to be purchased (e.g., a parking meter to pay with. Customers can register a new bank might display the code *145*980*122#). In this payment instrument in their mobile wallet by case, the customer initiates the payment request. requesting it from its issuer (through their branch, ATM, Internet banking, or telephone banking In any of these cases, the customer will receive channels) or by sending a message via short a message with the details of the transaction messaging service (SMS) with the keyword ALTA (description of the goods and the payment (subscribe) followed by the card number. Mobile amount), which the customer will need to authorize operators automatically register the mobile account by entering the personal identification number as a payment method (whether the customer is on (PIN) for the underlying bank payment instrument a prepay or postpay plan) when their customers (i.e., if using a debit card, it will be the same PIN use Mobipay for the first time. as the one used at the ATM) or by hitting OK if the transaction is charged to the operator bill (i.e., in The mobile phone acts merely as a payment this case there is no PIN). Buyer and seller will then initiator, because transaction processing is the receive messages confirming the transaction. responsibility of the selected payment issuer. The main Mobipay user interface is through Customers do not incur a charge using Mobipay. the Unstructured Supplementary Services Data Merchants pay the normal discount fees associated (USSD) mobile communications channel, which with the payment instrument used by the client works on most phones without requiring phone (e.g., a credit card), but there is no additional configuration. The USSD session can be triggered merchant fee payable to Mobipay if the payment directly by dialing a USSD number or, more occurs through its infrastructure. Mobipay pays commonly, by sending an SMS. This establishes an the customer's operator a fixed amount per interactive, real-time session between the mobile transaction (0.08 euro if billing against a card or phone and Mobipay's server. The user experience bank instrument and 0.15 euro if operator-billed), is a sequence of questions and answers. Mobipay for use of its communications channel. Mobipay has also developed interfaces using interactive charges a fixed amount to its shareholders to cover voice recognition (IVR), SMS (based on a system its costs. of keywords) or WAP.7 7 For a description and comparison of the USSD, IVR, SMS, and WAP mobile communication methods, see Mas and Kumar (2008), Box 1. 12 Use originally focused on mobile content (e.g., years later, there are only 400,000 registered--not ringtones); small on-the-road transactions, such necessarily active--users, amounting to less than as taxis (in the two major cities of Madrid and 1 percent of the population. And less than 2,000 Barcelona); public transport (buses in Málaga) and transactions are processed daily. parking ticketing (though not very widespread); and Internet purchases. In recent years, Mobipay This dismal performance can be explained by two has targeted the use of the platform for prepay main factors. First, Spain is highly penetrated with top ups. Customers need only send an SMS to a banking services and infrastructure, so Mobipay shortcode with the keyword recarga (load) and the struggled to open up a niche in the retail payments amount of the top up. market. Second, Mobipay did not have a marketing budget to promote its own service; it had to rely The Mobipay payment method is marketed by instead on promotion by its shareholders (who were operators and financial institutions and not by also its customers). These shareholders, in turn, did Mobipay itself. Customers are presented with the not see much benefit in promoting the Mobipay Mobipay logo to inform them of the availability brand, because they felt that their competitors of this payment method. While many retailers (whether the telecoms or the banks) would benefit have embraced payments through Mobipay, the equally from their marketing expenditures. As a largest department store chain (El Corte Inglés) result, Mobipay has languished in the absence of and the national airline (Iberia) have not. Mobipay effective marketing or a "killer application" that has remained fundamentally a tool for mobile- can raise public awareness of the service. enabling bank cards: these represent 90 percent of transactions, against only 10 percent being Mobipay's marketing morass is symptomatic of a operator-billed. larger misalignment of interests and experiences among Mobipay's diverse set of promoters, which Mobipay has recently been running a sequence has hampered its development. The telecoms of trials to try to broaden its market, including always saw their involvement as a concession the use of NFC-enabled phones to pay for public to the banks, which in fact were their major transport in Málaga, a parking ticketing application shareholders--a sort of shareholder tax.8 The in Madrid, a vending machine payment application telecoms' relatively tepid engagement is in part in Telefonica's corporate campus, and a betting because they were accustomed to margins of over payment application with a private company 50 percent in their prior content billing services called Bet And Win. Mobipay sees opportunities (e.g., ringtones/logos purchased using premium particularly in the transport area, with the support SMS), whereas the banks were happy to operate at of NFC-enabled devices--the space Octopus has much lower margins. Telecoms also saw the banks demonstrated. as overly conservative. The technical teams from banks and telecoms had very different mindsets, The market verdict and lessons learned used different terminology, and worried about different technical issues. And the fact that all Mobipay was trialed in mid-2002 in a small town and major telecoms and all major banks but one where launched nationally in late 2002. In less than a year, part of the Mobipay consortium meant that they it acquired 17,000 customers and 4,500 merchants faced no real competition--there was no sense of (2,800 online and 1,700 bricks and mortar). Six urgency at developing the market. 8 This is why all three major Spanish telecoms jumped at the opportunity of joining the Simpay consortium when it was conceived. That was going to be "their thing." 13 Mobile payments through able to use their mobile phones to transfer money credit cards: Moneta in Korea to other Moneta Cash accounts, including at POS to effect payments for goods. Transactions were What is Moneta capped at KRW 500,000 (US $500 approximately), but there were no limits on total value stored in SK Telecom (SKT) belongs to the third largest the card. The user proposition was based on the chaebol (conglomerate) in Korea. SKT controls convenience of being able to make payments around half the market for mobile telephony, from the mobile phone, without having to disclose with 20 million customers. It has developed a bank or credit card account details to payees. All comprehensive framework for mobile payments Moneta Cash payments were confirmed by SMS, (Moneta), mobile banking (Mbank), and mobile and customers got daily transaction records. commerce. Moneta was launched in November 2002 as a mobile wallet application that allowed Moneta Cash attained three million registered customers to make proximity (in-store) payments customers by 2004. Over time, tension arose with through several mechanisms. Moneta initially participating banks that increasingly saw SKT's supported a mobile cash payment product (Moneta m-cash accounts as an "invasion of an outsider Cash) and evolved toward a platform to support into their business domain" (Oh et al. 2006). In credit card payments through mobiles. 2004 sensitive customer data became accessible through the Internet and was used illicitly. Although Moneta developed in the face of stiff competition SKT claims it was not a technical problem but from other market players, especially from the rather inadequate care by certain customers in BankOn platform promoted by the powerful protecting their security information, Moneta Cash Kookmin Bank and supported by SKT's two mobile was discontinued following this incident. competitors (LG Telecom and KTF). The sweet spot: Mobile-enabling The origins: Moneta Cash credit card payments SKT pioneered mobile payments in Korea with a Unlike Mobile Cash, which gave customers a new mobile cash (m-cash) product, NeMo (Network type of account, Moneta itself was not a payment + Money), which was launched alongside nine instrument but rather a mobile wallet application major Korean banks in 2001 and was subsequently that allowed customers to pay using their credit rebranded Moneta Cash. Customers subscribing card over mobile terminals. It worked with a new to Moneta Cash got a virtual money account, with type of chip-embedded Moneta Card (a credit card their phone number acting as their account number. in a smartcard format). Moneta cards were launched M-cash account balances were maintained in an in September 2001 and were initially co-branded SKT server (i.e., not on the card), with SKT keeping by Visa and issued by five major domestic credit the float. Moneta Cash customers could use their card companies and banks. While SKT had made mobile phones to top up their m-cash account from plans to buy a credit card company to internalize linked bank accounts at participating banks. Both this aspect of the service, this never occurred. the m-cash account and the bank account were associated with the customer's phone number, In its original form, customers had to use specially and both accounts shared the same PIN. Once the designed mobile phones with a full-size smartcard Moneta Cash account was loaded, customers were reader into which they would insert their Moneta 14 Card when they wanted to make a payment against to change the chip inside their phone. Second, their credit card account. In March 2003, these Moneta customers were no longer restricted to phones were replaced with others that could house a narrow range of Moneta-enabled phones; they a subscriber identity module- (SIM) sized credit could use any standard third-generation phones. card smart chip (Moneta Chip) inside the phone, Third, this platform provided the ability to download so there was no longer a need to separately carry other mobile banking and public transportation and insert a Moneta Card to make a payment, and applications onto the same chip over the air, the phones could be smaller. Still, users needed rather than having to be preinstalled in each chip to have special-purpose phones,9 and customers (Bradford and Hayashi 2007). Thus, the Moneta wanting to pay with different credit cards would platform became the cornerstone in a broader set have to switch the smart chip inside the phone. of m-banking and m-commerce applications. Customers can use their cell phones incorporating Evolution beyond in-store payments the Moneta Chip in affiliated stores equipped with enhanced POS terminals. Merchants need Over time, SKT has expanded its Moneta payments to attach a dongle to their existing POS terminal; services from the merchant proximity payment this enables the terminal to read the Moneta Chip service (Moneta Card) to online payment services at close range using infrared or RFID technology. (August 2003) and mass transit payments (Moneta Waving the phone over the dongle triggers the Pass) in Seoul. In addition, it has used the payment payment. Additional PIN-based authentication platform to offer m-shopping, m-banking (Moneta through the carrier's SMS is required for higher Bank or MBank), and m-stock trading (Moneta value transactions. Moneta supports payments Stock Trading) services. in both online and offline mode. In stores not equipped with dongles, customers could continue MBank, launched in March 2004, is the result of to use a separate Moneta Card linked to the same a partnership of SKT with several smaller banks. credit card account. It was joined in September 2004 by the larger Kookmin Bank, which had earlier developed its own Since 2006, Visa and MasterCard have introduced m-banking platform. MBank uses multiapplication SIM-sized credit cards that work on their VisaWave chip technology to share the use of the Moneta and PayPass, respectively, proximity payment smart chip in the phone with the banks, so that platforms.10 As a result, merchant dongles became half of the chip is controlled by SKT to deliver its interoperable across alternative wireless service Moneta Card payments services, and the other half providers for the first time. is controlled by the bank to store relevant bank account information. In 2007, SKT partnered with Visa International and created a platform whereby subscribers The market verdict and lessons learned could download multiple credit card data and applications over the air to the standard SIM card By the end of 2003, SKT had placed some in their phones.11 This had three benefits for the 400,000 dongles with merchants. Moneta still had customer. First, customers could make payments only 40,000 users, despite the 400,000 Moneta- with any of their Visa credit cards without having compatible handsets available (Wallage 2003). 9 SKT's network uses the CDMA rather than GSM communications standard, and CDMA phones do not normally have SIM cards in them. Hence, use of the Moneta Chip required the user to have a special CDMA phone with a SIM card slot. 10 These are contactless smartcards that are fully compatible with the global EMV smartcard standard and can be read through insertion in traditional readers or wirelessly at close range (a few centimeters) using RFID. They give consumers a fast and convenient way to make payments, because no signature is required and there is no need to hand the card to the cashier. 11 SKT adopted WCDMA (which is part of the standard GSM technology evolution) as its third-generation network standard. The Universal SIM is a standards-based smart chip that is embedded in all WCDMA-enabled phones. 15 Moreover, only 21 percent of registered users really? In a world where people still carry cash and had made any purchase using their handsets. The hence wallets, what is the compelling customer greatest number of users came from the 30­39 benefit from transferring the credit card from age group (29 percent). By August 2005, SKT inside the wallet to inside the mobile phone? The had sold 4.9 million Moneta-enabled handsets, lack of a clear customer benefit was compounded a penetration of 20 percent of its customer base by usability issues, because the user interface (Beerfiles 2005). As of February 2007, there were design required customers to go through many 1.5 million registered users of Moneta services screens to make a simple payment (Bradford and in Korea (Payment News 2007), and 80 percent Hayashi 2007). of new third-generation-capable phones had the credit card functionality enabled. But according to Technology lock-in. In the absence of industry uncorroborated sources, use is very low, and the standards, market participants may chose to delay future of Moneta is uncertain. adopting new technologies until clear winners emerge. Many players were positioning themselves Moneta is fundamentally a new technology for with their own technologies to enable credit card paying from credit card accounts. With Moneta, payments over mobile phones: the two leading SKT sought to simply mobile-enable the preexisting telecoms, SK Telecom and KTF, were promoting payment ecosystem. By not introducing a new their own standards; Harex InfoTech, a start-up, payment instrument, SKT did not need to convince offered its own infrared-based m-payment service users about the safety and soundness of a new called ZOOP in parts of Seoul; and some established type of money (unlike Mondex or Octopus, which credit card companies wanted to develop their positioned a new prepaid account). Also, by own card-based m-payment solutions. keeping the payment instrument with bank issuers rather than taking the issuance role itself, SKT did · The solution required installing proprietary not need to itself assume any collection duties merchant POS readers, which for many years (unlike Simpay, which favored operator billing). were not interoperable among rival systems. Retailers faced the need to have to deploy Yet despite this apparent "simplification" of the multiple card-accepting devices, which would mobile money innovation, Moneta has experienced add cost and complexity to their operations and relatively low levels of use in a population that ultimately eliminate the efficiency benefits from is among the most tech-savvy in the world. adopting e-payments. Retailers resisted investing Driving adoption of a new way of presenting and in the new equipment (dongles) necessary to exchanging credit card information has proven process Moneta transactions before demand for difficult for three main reasons. such services was well proven. · Similarly, handset vendors were slow to respond Utility. From the customer point of view, it is not in developing the special-purpose Moneta apparent what problem Moneta is addressing capabilities (with the contactless dual chip) until relative to standard credit card payment they saw the market as fully developed (Wallage arrangements. At its most basic, Moneta simply 2003). Korean handset manufacturers have physically combines the mobile phone with the always been keen to showcase new technology. credit card. How much of an advantage is that However in recent years they have become more 16 successful in global handset supply and thus the and Telefonica Moviles. Two smaller national ability to have a product that is globally more operators, Amena and Proximus, joined the applicable is influencing their engagement in project in early 2005. Its mission was to develop local initiatives. and operate a pan-European payments system for mobile phones, focused on micropayments (less Distrust from financial institutions. The than 10 euros). Despite the company's tagline-- development of m-payments was hampered by "pay for stuff with your mobile"--the system very public bickering between leading banks and was also designed to allow purchases from PCs telecoms and the consortia they formed. connected to the Internet. · There was little alignment in the business Before Simpay, mobile commerce was largely models. The banking and credit card industries driven by purchases of low-value digital content-- were not supportive because the mobile carriers ringtones, obtaining sports results--from demanded a large share of the transaction operators' proprietary portals ("walled gardens"), revenue (Bradford and Hayashi 2007). At the and customer billing was through premium-rate same time, it is reported that SKT wanted credit SMS text messages between the customer and card fees reduced to 1 percent (from the normal the merchant. With the advent of new third- 2­3 percent) on credit card transactions taking generation networks, operators saw much larger place through its network to stimulate use mobile commerce opportunities, such as music (Wireless Week 2002). downloads, Java games, video streaming, and TV · The typical bank­telecom disputes over customer over mobile, from data services. ownership flared up. The single Moneta Chip houses SKT subscriber data and the customer's Through Simpay, operators sought to make it credit card and possibly bank account easier for their customers to buy digital content information. Banks remained concerned that from any source, whether on- or off-portal, using SKT's control over the Moneta Chip would allow any connected device (mobile phone or PC), from SKT to control what services were proposed to any country covered by the operator consortium. their customers. It took until 2004 for SKT and Retailers would be drawn by the fact that on joining Kookmin Bank (the leader of a rival scheme) to Simpay and interfacing once into its systems, they determine how to collaborate.12 could then gain access to all the customers of Simpay's member operators, which amounted to A multioperator platform 280 million at the time. Customers would be able for mobile operator billing: to conveniently purchase digital content directly Simpay in Europe from a broader range of retailers, who would advertise the availability of the Simpay payment What is Simpay method as a trusted payment brand. Simpay (initially called the Mobile Payments Simpay is a mobile operator-centric model, based Services Association) was launched in February on operator billing mechanisms. To the extent that 2003 by a consortium of the four leading European it targeted small transactions, it could be seen as mobile operators: Orange, Vodafone, T-Mobile, complementary to credit and debit cards and could 12 Interestingly, Oh et al. (2006) write: "Although Kookmin Bank joined [rival SKT's] MBank, Kookmin was considered as the winner because it successfully prevented SK Telecom from the rights to control customer information." 17 operate within the requirements of the European Thus, the Simpay initiative sought to create a Union's eMoney directive. de facto standard controlled entirely by mobile operators. It envisioned a standardized set of However, it represented a potential longer technical interfaces across the billing platforms term threat to credit cards if it came to replace of all member networks and promised to reduce credit card use among some customers. On the the number of contractual relationships needed other hand, Simpay considered the possibility between operators and merchant third parties. in the longer term to offer customer "wallet" A merchant could plug in through a merchant functionality, which would enable the customer acquirer and sell its products to the customer of any to choose among payment methods (e.g., on the participating network. And the Simpay logo would operator bill, on credit card, or on a stored-value be deployed widely to inform customers about the account). availability of the mobile payment channel. Subsequently, the platform would also support Cross-border transactions (e.g., where the user is payment for larger purchases undertaken on the purchasing content from a provider in a different mobile or PC (e.g., theater tickets), as well as country) were supposed to be firm at time of physical (in-store) transactions. Some operators saw transaction. Simpay would negotiate exchange Simpay as ultimately becoming a cash alternative. rates ahead of time with the settlement bank and apply it to all transactions within a particular day. How it works This offered more pricing certainty to customers than they typically get when they pay with credit Under the Simpay scheme, mobile purchases cards, because in this case the customer does not are debited from the mobile user's account-- know the cost in his home currency at the time of whether prepaid or postpaid--with his operator. transaction. The operator thus authenticates, provides the mobile transacting channel to, and bills the user. Simpay was also intended to enable operators Transactions with merchants (retailers and content to profile customers for marketing and service providers) are aggregated through merchant purposes. By incorporating customer payment acquirers--essentially intermediaries who would data into their customer relationship management driveadoptionofthetechnicalplatformbyindividual systems, mobile operators saw an opportunity merchants and who would channel payments to develop a more effective customer marketing between Simpay and individual merchants. Simpay channel and create more customer "stickiness." stands between the various mobile operators and the various merchant acquirers, providing payment The demise of Simpay authorization, clearing, and settlement services between them. Operators can also act as merchant Simpay selected four major vendors to develop, acquirers--as when the user purchases content test, and run the system: Valista, Encorus, Privnet, from the operator's own content portal (or walled and Qpass. The initial plan was to offer the service garden)--but they also allow users to purchase in 20 countries by 2004, but many delays ensued. content from third-party merchant acquirers. Simpay was hampered by strategic and operational difficulties. Despite the common vision at the 18 beginning, the sponsors were unable to agree working level, there was even more tension between on what types of mobile payments Simpay would the operators on one side and Simpay staff on the offer. T-Mobile wanted to incorporate contactless other. Simpay had recruited a significant number payment chips within the mobile phone casings to of staff with a banking or payments background, take on proximity payments, whereas Vodafone which clashed with operators' mindset. wanted to concentrate on purchases of digital content. The market (non)verdict and lessons learned These misalignments stemmed in part from the fact that Simpay's business model was squeezed Divergences among the founding members led by two operator concerns. First, they did not want T-Mobile to withdraw from Simpay in June 2005, to lose the high margins (60 percent or so) on and Simpay's activities were discontinued shortly existing premium SMS sales, mainly for ringtones thereafter. Simpay was never launched. and logos. Second, many of the participants in Simpay had previously established their own Unlike Mondex, Simpay was, at least at its mobile Internet payment services, which in fact conception, not in the business of replacing remain in place today, such as Mobipay for the cash. And like Octopus, it tried to concentrate Spanish operators and Vodafone UK's M-pay.13 on small payments where cash was impractical The common goal was to open up new mobile and in a vending environment it could tightly commerce streams, but it was unclear what control. Indeed, by focusing on remote payments such streams actually were. Although there for digital content, Simpay would offer a valuable was a genuine interest by the operators in the new payment mechanism to its customers while consortium to harmonize their platforms, some playing to the strengths of its member operators. member operators were pursuing their own related And by basing payments on the billing platforms research and development activities in parallel, as if of its member operators, it also avoided the bank­ they were hedging their bets. telecom partnership tensions that sapped Mobipay and restricted the potential of Moneta. Thus, It appears that the complexity of the task had been Simpay was charting a new path that in many ways considerably underestimated. The solution was incorporates the lessons from the experiences that deemed by many to be out of proportion for the we have reviewed so far in this paper. very low volumes of paid mobile services at the time, beyond ringtones and logos. Some people However, as development work proceeded, questioned whether it was really so important to member operators started diverging on their create a pan-European payment roaming network, overall product vision. Moreover, Simpay's task was at least in the first instance, which inevitably forced complicated by the multicountry, multioperator a solution that was not optimized to "fit" the nature of the platform it needed to build. Ultimately specific needs of any given market. it was not able to maintain strategic alignment among its member operators for several reasons: Over time, as strategic divergences became more marked, the operator working groups creating the · From a product marketing point of view, it specifications lost momentum. Many were there was never clear what key benefits were being to ensure that their competitors didn't get ahead pursued. Many questioned whether international rather than to pursue a common objective. At the interoperability was the right focus, because 13 M-pay has evolved to support the U.K.-wide "payforit" service, which provides billing to mobile accounts for any service designated on the Web as being "payforit-enabled." This service is now available across all licensed U.K. mobile operators. 19 much content is of a local nature. And if retailers Osaifu-Keitai is based on a FeliCa card embedded in were to be nationally aligned with operators, mobile phones (the same that was used by Octopus it was not clear what advantage the platform in Hong Kong). The FeliCa card was developed would have over existing operator-specific by Sony, and the chipsets are commercialized premium SMS-based billing methods. It is always by FeliCa Networks, which is controlled by Sony a challenge to build harmonized products that with a minority stake (38 percent) by DCM. The do not lose sight of customer wants and needs. FeliCa card includes an integrated nonvolatile · From a business model point of view, many memory that makes it possible for mobile devices operators were loath to relinquish the high to contain multiple forms of data, including margins on existing ringtone/logo sales through personal identification as well as up to 10 service- premium SMS and, hence, relegated Simpay to specific applications. The FeliCa card also includes be the payment mechanism for future, though a wireless communication chip, enabling data as yet unavailable, mobile Internet content exchange when it is held adjacent to compatible sales. The fear of cannibalizing existing revenues readers/writers via RFID. This allows consumers undermined Simpay's business case. to use their phones as a substitute for cash and · From a governance point of view, a consortium cards at vending machines and merchants' POS; in of such major industry players is very difficult to addition it facilitates remote payments. maintain, even within the apparent homogeneity of an operator club. Despite initial alignment, the Osaifu-Keitai is a device-based mobile payments true motivation and strength of interest of the solution, supporting both proximity payments in partners changed as the project unfolded. Some shops that have a FeliCa chip reader and remote operators had more interest in checking the (online) payments. Although the FeliCa-based interests of their competitors than in pursuing wallet platform was developed by DCM, since the common good. 2005 it has been supported by the two other major mobile phone operators in Japan, making it the Without clear product marketing vision coming de facto standard mobile payment platform in from the member operators, Simpay ended up Japan. As a result, FeliCa readers (though not the developing technical plans that were unrealistic. payment services themselves) work across mobile Each operator had special technical interface operators. requirements, which led to an overly complex technical architecture. Osaifu-Keitai is based on a wallet application on the FeliCa chip, which supports multiple payments An operator-centric instruments (Williams 2008): transactional ecosystem: Osaifu-Keitai in Japan · Credit cards. DCM developed a proprietary platform, iD, for storing e-credit cards on FeliCa- What is Osaifu-Keitai enabled handsets (analogous in its functionality to the Moneta Chip system described previously). DCM is Japan's leading mobile communications It is open to the cards of other issuers, in addition operator, with 53 million subscribers as of March to DCM's own DCMX credit cards.14 Payments 2008, representing over half of Japan's cellular against DCMX cards are billed together with the market. It launched a mobile wallet service, Osaifu- user's monthly DCM phone charges. Keitai (meaning wallet-mobile phone) in July 2004. 14 In Japan, credit card issuers need not be banks. DCM purchased a 33 percent stake in Sumitomo Mitsui Card. 20 · Operator billing. Small expenditures can be recognized the challenge of convincing retail chains charged directly to the phone bill. This is called to accept their mobile money scheme. They were DCMX mini, and it has a 10,000-yen (US$94) deliberate in stressing to these agents the ways in credit limit. Big spenders can apply for more which Osaifu-Keitai would increase both customer credit and use it just like a regular credit card. convenience and value for their businesses through · Prepaid e-money. DCM entered a joint venture (1) process speed, (2) versatility, and (3) security called bitWallet with Sony and some financial (Jenkins 2008). institutions. bitWallet offers a prepaid e-money service called Edy.15 Customers can fund their However, DCM appears to have retained a tight Edy accounts with cash or credit cards or from degree of control over the payment options Internet/mobile banking accounts. The Edy underpinning Osaifu-Keitai. The credit card wallet is limited to small-value settlements (up payment system iD was proprietary to DCM, to 3000 yen) and can hold up to 50,000 yen in requiring merchants to install iD-compatible stored value. As of February 2008, it is accepted readers. It has stated that it will seek to collaborate in more than 71,000 convenience stores, more closely with banks and card companies to bookshops, and coffee chains and at vending further expand Japan's mobile payments market machines. The Edy network handles close to a (Market Platform Dynamics 2006). In 2007, DCM, million transactions per day. East Japan Railways, JCB, and bitWallet agreed to · Transport ticketing. East Japan Railways (the make their readers interoperable (Bradford and third partner with a 5 percent stake in FeliCa Hayashi 2007). Networks) developed Mobile Suica, which allows customers to buy railway fares from a prepaid The market verdict and lessons learned account. Like Edy, this can be funded from cash, credit cards, or Internet/mobile accounts. As of January 2006, there were over 10 million · Postpay services. Beginning in 2005, JCB and subscribers of Osaifu-Keitai with compatible AEON Credit Services have released QuicPay, a handsets. As of 2008, there are more than 29 postpay service. million subscribers (NTT DoCoMo 2008). The mobile wallet application is accepted at more than DCM's core belief--one that it developed based 640,000 stores (Contactless News 2008). on the success of the i-mode content platform in Japan-- is that use of payment services will be Osaifu-Keitai has achieved good recognition in the pulled by the presentation of a range of services marketplace, but usage is lagging. A survey of to the customer. Therefore, DCM has developed 1,093 mobile users in Japan in June 2007 showed a very structured approach to enticing partners that while only 40 percent had Osaifu-Keitai-enabled onto its content and payment platforms and to phones, twice as many knew about the service. Of promote further service innovation. The service those with the capability to use Osaifu-Keitai, only platform now supports a variety of uses beyond 30 percent used it "sometimes" or "often."16 e-money, including airplane tickets, membership cards, and house entry keys. Users can download Though Osaifu-Keitai is by no means a proven multiple applications over the air onto the FeliCa success, it does show significant promise. What card to support each of these services. DCM also sets it apart from all the other initiatives we have 15 DCM has a 14 percent stake in bitWallet. 16 The survey was conducted online in June 2007 by Goo Research, and reported in http://whatjapanthinks.com/tag/osaifu-keitai/. 21 reviewed is the sheer level of dominance DCM these markets are already so "banked." There is no exerts over the mobile market (with an impressive need to pay with my mobile phone, when I have a 80 million customers) and the ecosystem around debit or credit card in my wallet. There is no need it. DCM is able to heavily influence mobile handset to totally eliminate cash when there are ATMs design because of its close relationship with on every corner. So we can view these unsettling manufacturers, through which it can streamline experiences in a different light: a quiet technology the customer experience; it purchased a bank battle has occurred for e-payments, and the debit and transaction processing company to drive card has won. It is not a case of repeated failures economies of scope; it provided substantial funds to make payments electronic, but rather it's a case as an incentive to merchants to purchase the of a singular success: the debit card. From the necessary POS terminals; and through its base developing world perspective, however, many of of developers and development partners it was these services are so out of reach for people that able to foster substantial innovation around new mobile phones and smartcards may have a role to platforms, such as Osaifu-Keitai. The sheer muscle play. It is therefore important to understand the that DCM brings is unparalleled. This approach stark economics these schemes must defeat to would not be replicable in Europe, where markets succeed. are more fragmented on both the telecom and banking sides. The problem with new payment devices Learning from these experiences The experience has amply demonstrated that new forms of e-payments beyond debit cards The above experiences show very limited success cannot be simply swept in by the spread of with electronic or mobile money schemes digital technologies, such as smartcards and beyond debit cards. The lack of a compelling mobile devices. They need to deliver economic user experience and lingering security concerns advantages to both buyers and sellers, and they remain barriers to customer adoption. Transport need to be in every way at least as reliable and applications are driving most of the usage today convenient to use as the forms of payment they and appear to be the main focus of new business are displacing. There is still no money scheme that development. Nontransport-related service leverages the data entry, storage, processing, and introductions appear to be primarily aimed at communications capabilities of mobile phones showcasing the technology in key markets, such (or, their more limited counterparts, smartcards) as Japan and Korea, but with as yet limited in a way that can be expected to be universally commercial rationale. Although customer data adopted and replicable. are very hard to come by, penetration rates are reasonably high in some schemes but usage is Why is it so difficult? Why is there so much market very low across the board.17 Box 1 provides an inertia supporting old ways of doing things, in the (admittedly crude) indication of just how much in face of significant technical innovation? There are the news these schemes now are. three broad categories of reasons: The partial successes and many failures of these · Those relating to the strict comparison, from initiatives in developed countries are mainly because the customer's point of view, of the use features 17 Van Hove (2005) shows data on eight e-money schemes in Europe, including number of cards issued, number of acceptance points (terminals), frequency of use, and float. 22 Box 1: How Current Are They? We wanted to get an indication of how "hot" the documents behind these top 10 search results by various e-money schemes are today--are they still year of publication. The results are shown in the table much discussed and analyzed, or are they yesterday's below. news? To get a simple indication of that, we conducted Google searches on each of the schemes reviewed Osaifu-Keitai ranks as the most talked about, in this paper (the exact key words are shown in the followed by Octopus, and then Moneta. Mobipay first column of the table below and were chosen to is failing to generate much excitement. And be as broad as possible without creating ambiguity predictably, the moribund Mondex and the defunct with other possible uses of the terms). For each Simpay schemes have not received any attention scheme, we picked out the first (and, presumably, in recent years. While a Google search may seem most relevant and popular) 10 search results with like a crude way of measuring the relative success or meaningful commentary or analysis on each scheme. failure of e-money schemes, this proxy does confirm We ignored search results relating to the company's the story we've told about the rise and fall of various own Web site or press releases. We then dated the e-money schemes. 96 97 98 99 00 01 02 03 04 05 06 07 08 Mondex Smart Card 2 2 3 2 1 Octopus Card 1 2 3 4 Mobipay 1 2 1 3 1 1 1 Moneta Korea 1 2 1 2 3 1 Simpay 1 2 7 Osaifu-Keitai 1 1 3 5 Total By Year 2 2 3 2 0 1 3 4 7 10 5 10 11 Note: Seaches conducted on google.com during 27­28 August 2008. among cash, debit cards, and the new forms For all the problems associated with the physical of e-money and payment systems. Users will embodiment of cash, it has some unique hold the burden of proof on the new payment characteristics that make it very useful: it is familiar devices to demonstrate superiority in at least to everyone through centuries of use, easy to some applications. value (counting), quick to exchange (at least for · Those relating to the difficulty for the public to low values), guaranteed to be anonymous (no assess the security and establish trust with the traceability), sturdy (withstands harsh handling), new system. universally accepted (no basis for exclusion, de · Those relating to the difficulty of supplanting jure as well as de facto), divisible (you don't need established payment mechanisms and standards to take all of it with you), and incurs no marginal due to the microeconomic characteristics of transaction cost (at least once you have the cash payments markets. Even if a new payment device in hand). Van Hove (2007) stresses that people is intrinsically clearly superior, that by itself will value the underlying product purchased and not not guarantee its success in the market. the payment service itself, and so these attributes 23 of cash need to be viewed individually and point of view, there is a premise underpinning collectively as hygiene factors rather than merely many of the schemes we have reviewed, whether as desirable features that can be traded off against they were based on new forms of e-money or on each other. enabling debit or credit card transactions, that people benefit from consolidating the contents The e-money schemes that have been tried may be of their pocket--carrying a card instead of a full superior to cash and debit cards in many ways, but wallet and a mobile phone instead of a card. That each is inferior in at least one way that has proven may not be the case, and in fact it is possible that to be important to customers: people want money to be embodied in distinct tokens (coins, bills, cards) rather than merging into · Card-based e-money products require some fuller service devices, such as mobile phones. familiarity and comfort with technology. They need to be expressly loaded. Their value cannot There is a particular challenge in overcoming be readily observed by customers. They are public concerns on the security and reliability of less sturdy than cash and expose the user to new money schemes. People respond better to malfunction, in which case the user stands to known risks (probability of being robbed on the lose the stored value. Making payments depends street or of a stash of cash being found) than to on finding a card-reading device, which greatly unknown ones (e-security breaches). They are limits the ability to make direct P2P payments. also more ready to accept loss of value when it · Network-based e-money products working online happens because of their own actions or inaction are safer (because the value is not stored in the (e.g., distraction leading to loss of cash) than card itself), but financial intermediaries holding because of circumstances beyond their control the accounts of payer and payee need to be (e.g., smartcard malfunction). As a result, it is very involved. As a result, they are slower to transact difficult to reassure customers about new payment than cash (because of the need for PIN-based devices when they lack a track record and require security and real-time transaction authorization higher levels of technical sophistication. These with the issuing institution), incur a marginal concerns get in the way of people even trying to communications cost, and may be traceable. But learn about and experiment with new schemes. then they work analogously to a debit card, and A recent poll across many developed countries it is not clear why customers would choose to suggested that 60 percent of people continue to fragment their liquidity between bank account "not trust their mobile device to provide a secure and network-based e-money. transaction," and only 10 percent "are likely to do m-commerce."18 These problems are magnified by the fact that the costs and benefits of cash and alternative payment Even if there were a new payments device that schemes are not distributed evenly across all players. was decidedly superior from a user's point of view, From the merchant side, it is common for smaller there are several characteristics of the e-payments merchants to prefer cash despite the theft and market--common to many network-based fraud concerns, and it is only the larger merchants markets--that militate against rapid adoption and who want to improve efficiency by adopting favor incumbent solutions (such as cash or debit e-payments (Jones 2008). From the consumer's cards): 18 Telecom Asia (2008), quoting a study by Unisys. 24 · Network economies: the value of the payments unique advantage over cash have met more network grows with the number of economic (though still limited) success than general agents using it. Getting early adopters is difficult e-payment schemes. Clear opportunities exist to when there are few other users to transact address purchases at unattended POS (such as with, and more established schemes become the fixed denomination, small-value transactions unassailable. at ticketing, parking, or vending machines · Economies of scale: e-payments are enabled by Octopus or Suica (powered by characterized by large fixed costs (merchant Osaifu-Keitai) and remote purchases over the terminals, value loading machines, servers) and Internet or mobile phones (such as mobile very small marginal costs. This creates significant content downloads enabled by Mobipay or barriers to entry and delays the point at which Osaifu-Keitai). newer solutions can be justified on a business · Interoperability and industry consortia are case basis. a double-edged sword. Schemes based · Two-sided market: electronic forms of payment on interoperability, such as Simpay, can be need to deal with penetrating two sides of the subjected to large coordination problems among market--customers and merchants--and both the various players. They also risk being pulled need to be cracked at the same time, which into multiple strategic directions, with adverse creates a familiar chicken-and-egg problem. The impacts on technical complexity, service cost, problem is particularly acute when the two types and quality of the customer experience. Yet of users need to incur specific costs (e.g., for proprietary schemes, such as SKT's, are much payment terminals) to participate in the market. harder to get adopted in the marketplace, because they require specific investments in This triple whammy means most new payment their own handsets and merchant terminals. schemes will need a significant amount of time They also can lead to market paralysis as market to establish themselves and grow, even if they participants defer investment decisions until a offer distinct customer benefits. But while they do clear winner emerges in the marketplace. not supplant the old schemes, their value will be · Customer convenience and in particular speed undermined: why adopt e-money if I will still need of transaction is essential. The more successful to carry cash around for some purchases? Hence, schemes--Octopus, Osaifu-Keitai, Moneta-- both store customers and merchants may have are all based on devices with short-range limited incentive to use the new payment method communications (RFID or NFC) that work with while it is being rolled out in the market, even contactless card readers and minimize the data though using it is in their long-term interest. entry requirement on users. · Payment enablement needs to be conceived But if you must try it... as a platform or ecosystem, not as a service. Providers have to not so much sell as orchestrate Our review of different experiences has highlighted the creation of a new market. Driving adoption specific lessons: requires complex multi-party gaming strategies, and at the very least takes time. As DCM has · Schemes that have developed within captive, shown, these ecosystems need to develop in niche markets where they can exploit some a fairly open, collaborative fashion, but at the 25 same time they need to be forcefully led. Few References players can be expected to have the power and the skill to bring this about. Beerfiles. 2005. "Wallet Phones for Asia." Blogpost, 15 September. http://www.beerfiles.com.au/ Many observers believe that the most promising content/blogsection/11/63/. area for e-payments in developed countries at least in the near future is ticketing. The above Birch, Dave. 2002. "Purse Wars--What does the list of lessons and the Octopus case study shows front line look like?" http://www.arraydev.com/ why. Octopus exploits a niche where cash is not commerce/jibc/9806-02.htm. so convenient: at unmanned ticket machines. The backer of Octopus was able to drive fast Bolt, Wilko, and David Humphrey. 2007. "Payment take-up by customers by adjusting ticket pricing Network Scale Economies , SEPA, and Cash to favor the new pricing scheme. At the same Replacement." Research Department Working time, the backer of Octopus controlled a very Paper No. 07-32. Philadelphia: Federal Reserve sizable network of acceptance terminals, which it Bank of Philadelphia, July. http://papers.ssrn.com/ was able to convert to the new payment scheme sol3/papers.cfm?abstract_id=1077197. overnight. This double approach, "easing pain" by adopting smartcards and removing the need Bradford, Terry, and Fumiko Hayashi. 2007. to deal with ticket machines and "applying pain" "Complex Landscapes: Mobile Payments in Japan, on cash purchases through higher prices, provided South Korea, and the United States." Payments powerful incentives that drove very rapid adoption. System Research Briefing. Kansas City: Federal Once people become aware of Octopus and are Reserve Bank of Kansas City, September. comfortable using it, its usage broadens naturally beyond the original purposes. Brits, Hans, and Carlo Winder. 2005. "Payments Are No Free Lunch." http://www.cepr.org/meets/ In developing countries mobile payments and wkcn/1/1627/papers/Winder.pdf. e-money may have a bigger role to play given the paucity of banking infrastructure and alternative Chua, Thian Yee. 2006. "Building Unique Service payment channels for people without access to bank Convergence Ecosystem." Singapore: Cassis accounts. The market opportunity may be larger, International, April. but schemes in developing countries will face similar challenges in establishing critical mass. Clarke, Roger. 1996. "The Mondex Value-Card Scheme: A Mid-Term Report." Version 2.0. http://www.anu.edu.au/people/Roger.Clarke/EC/ Mondex.html. Contactless News. 2008. "Starhub to Pilot Mobile Wallet Based on Japan's Osaifu-Keitai Service." 18 June. www.contactlessnews.com/2008/06/18/ starhub-to-pilot-mobile-wallet-based-on-japans- osaifu-keitai-service. 26 De Jong, Iddo. 2006. "Prepaid Cards: A Discussion Papadopoulos, Georgios. 2007. "Electronic Money Paper." European Central Bank, Directorate and the Possibility of a Cashless Society." Working General on Payments Systems and Market Paper 18.02.2007. Rotterdam: EIPE. http://papers. Infrastructure. http://epso.intrasoft.lu/papers/ ssrn.com/sol3/papers.cfm?abstract_id=982781. De_%20JONG_Prepaid_Cards.pdf. Payment News. 2007. "Visa, SK Telecom Announce Ives, Blake, and Michael Earl. 1997. "Mondex Plans to Launch Mobile Payments in South International: Reengineering Money." London: Korea." 8 February. http://www.paymentsnews. London Business School. http://mis.uis.edu/ com/2007/02/visa_sk_telecom.html. ecomm2/mondex_case/mondex.html. Simpay White Paper. 2004. "Simpay--Pay for Stuff Jenkins, Beth. 2008. "Developing Mobile Money with Your Mobile." Ecosystems." IFC and Harvard Kennedy School. Siu, Lucia. 2002. "Octopus and Mondex: The Social Jones, Peter. 2008. "Europe Set to Lose its War Shaping of Money, Technology and Consensus." on Cash." London: PSE Consulting, March. http:// http://www.sociology.ed.ac.uk/finance/Papers/ www.psel.co.uk/pdf/articles/general/europe_set_ Siu_Octopus.pdf. to_lose_its_war_on_cash_mar08.pdf. Stadler, Felix. 1998. "Mondex: Early Problems of Market Platform Dynamics. 2006. "The Global Implementation." E-Money, Vol. 1, No. 7. http:// Adoption and Evolution of eWallets." MPD felix.openflows.com/html/mondexguelph.html. Dialogue Series: Payment Industry Futures, 11 January. http://www.marketplatforms.com/MPD/ ------. 2002. "Failures and Successes: Notes corporate/newsandevents/Events/Other%20 on the Development of Electronic Cash." The Events/. Information Society, Vol. 18, No. 3. http://felix. openflows.com/html/ANT_Mondex.html. Mobipay. 2008. Dossier Informativo. http://www. Mobipay.es/pdf/Dossier_Informativo_Mobipay. Telecom Asia. 2008. "Consumers Still Wary of pdf M-Commerce." 28 August, p. 25. National Bank of Belgium. 2006. "Costs, Van Hove, Leo. 2005. "Electronic purses, Advantages and Drawbacks of the Various Means interoperability, and the Internet." In Fox, M. A., of Payment." Economic Review (June): 41­47. ed. Special Issue on Internet banking, e-money, and Internet gift economies. First Monday, Special NTT DoCoMo. 2008. "Factbook PowerPoint Issue #3, 5 December. http://firstmonday.org/ Presentation." July. issues/special10_12/. NTT DoCoMo. www.nttDoCoMo.com ------. 2007. "Central Banks and Payment Instruments: A Serious Case of Schizophrenia." Oh, S., H. Lee, S. Kurnia, and R. Johnston. 2005. Communications & Strategies, no. 66, 2nd "Competition and Collaboration in Mobile Banking: quarter, p. 19. http://papers.ssrn.com/sol3/papers. A Stakeholder Analysis." Hong Kong Mobility cfm?abstract_id=1005111. Roundtable. Hong Kong: HKUST, 2­3 June. 27 Wadsworth, Jim. 2004. "Simpay: Right Product, Williams, Martyn. 2008. "Only in Japan: The Best Right Time." Presentation to the Digital Money Technologies You Can't Buy." PC World, 11 Forum, March. February. www.pcworld.com/article/142120-2/ only_in_japan_the_best_technologies_you_cant_ Wallage, Steve. 2003. "The Far East Mobile buy.html. Payment Race." TheFeature.com, 27 November. www.thefeaturearchives.com/topic/Business/The_ Wireless Week. 2002. "Korea's mPayment Plans Far_East_Mobile_Payment_Race.html. Hit By Conflicts." 14 October. www.epaynews. com/index.cgi?survey=&keywords=SKT&optional A Week in Wireless. 2008. "New Broom." No. 323, =&subject=&location=&ref=keyword&f=view&id 5 September. =1034594920622215212&block. No. 51 December 2008 Please share this Focus Note with your colleagues or request extra copies of this paper or others in this series. CGAP welcomes your comments on this paper. All CGAP publications are available on the CGAP Web site at www.cgap.org. CGAP 1818 H Street, NW MSN P3-300 Washington, DC 20433 USA Tel: 202-473-9594 Fax: 202-522-3744 Email: cgap@worldbank.org © CGAP, 2008 Ignacio Mas is a senior adviser and Sarah Rotman is an associate providing very useful insights and information and Sangwoo Yu, microfinance analyst, both in CGAP's Technology Program. Hyesoon Kim, and Jaesic Jeon of SK Telecom for providing facts The authors wish to thank David Bridge and David del Ser for and comments on Moneta. CGAP publications are frequently cited in other works. The suggested citation for this paper is as follows: Mas, Ignacio, and Sarah Rotman. 2008. "Going Cashless at the Point of Sale: Hits and Misses in Developed Countries." Focus Note 51. Washington, D.C.: CGAP, December.