Mobile Payments: At the Tipping Point
SEE LAST PAGE OF THIS REPORT Howard Mason
FOR IMPORTANT DISCLOSURES 203.901.1635
August 5, 2013
Mobile Payments: At the Tipping Point
- Last Wednesday’s announcement of a national rollout for Isis, a near-field communication (“NFC”) system for mobile payments, will catalyze adoption of the NFC mobile payments infrastructure and usage of “tap and pay” digital wallets. Partner carriers (Verizon, AT&T, and T-Mobile) and banks (JPM and COF) have resolved the dispute over access to payment account information on the SIM card which has hobbled Google Wallet.
- By 2017, we expect substantially all US smartphones to ship with NFC capabilities and 80% of merchants to have NFC-enabled terminals. Presently ~15% of smartphones are NFC-enabled and only 5% of merchants are NFC-ready (although 25% of the largest 100 retailers have either deployed, or are deploying, the technology).
- We agree with estimates from e-Marketer magazine that phone-based payments at point-of-sale will increase from $1 billion in 2013 to $10 billion by 2015 to over $50 billion by 2017.
- As NFC adoption becomes increasingly evident, investment in image-based technologies for mobile payments relying on barcodes, including the MCX system by a consortium of merchants including WalMart and Target, will decline.
- We expect Bank of America and USB to join the Isis issuing banks.
- Square and other closed-loop mobile payments solutions (in which customers pre-register with merchants so that there is no need for information exchange between a smartphone and merchant terminal) have a business case if they can lever access to merchant systems to integrate payments acceptance with broader customer relationship management.
- OpenTable is a good example where payments will be integrated into the broader computerization of the customer relationship from electronic booking through checkout.
- Given the announced partnership for mobile payments with Discover, PayPal can follow its current closed-loop approach or embrace NFC technology using Discover’s contactless Zip acceptance brand; a hybrid approach is likely.
- Merchants will seek to interpose proprietary cards in digital wallets between Visa accounts and checkout, as Starbucks has done. These will allow them to implement rewards programs based on customer shopping behavior (possibly with location- and/or time-based notifications) without sharing data with Visa.
- Longer run, merchants will use rewards to incentivize customers to fund the interposed cards with ACH-accessed bank accounts rather than Visa or MasterCard accounts; if successful, this can blunt and even reverse the cycle of increasing credit interchange by engaging consumers in the brand of the accepting merchant over the issuing bank.
Section I: Isis Will Establish NFC as the Mobile Payments Standard
Use of Mobile Payments Has Disappointed
There is a gap between the promise of mobile payments and the practice. Consumers are aware of mobile payments with a Jan 2013 survey by research firm Chadwick Martin Bailey (“CMB”) finding 50% of smartphone users familiar with the idea; and users like the benefits according to Jan 2013 by PriceWaterhouseCoopers which found 80% of users liked mobile payments for the opportunity to save money and nearly 60% for its convenience and ability to save time.
However, usage is low at only 16% of US smartphone users among the group surveyed in January by CMB and the industry is disappointed as reflected in reduced estimates for mobile payment volumes. For example, just last month, eMarketer magazine cut in half its US estimate for phone-based payments at point-of-sale (“POS”) to $10bn in 2015 from a prior estimate of over $20 billion; the estimate for this year is $1bn (see Exhibit 1).
Exhibit 1: Forecast Mobile Payment Volumes
Delayed Rollout of Mobile Payments Infrastructure
The gap between consumer interest in mobile payments and consumer usage arises because of delays to the roll-out of a mobile infrastructure and, specifically, the near-field communications (“NFC”) infrastructure allowing customers to use radio frequency identification (“RFID”) and payment authentication. NFC technology has been validated as an industry standard for contactless cards, such as Exxon Speedpass which is accepted at two-thirds of Exxon’s 16,000 gas stations in the US, and involves a customer waving or tapping a card or key fob over a card reader equipped with a radio sensor.
Note: This “tap and pay” procedure is different from a traditional credit card where identity information is stored on a magnetic stripe which is “swiped” at a card reader (or, following Square’s breakthrough idea of converting information encoded on a “mag” stripe to an audio signal capable of passing through an audio jack, a dongle attached to a smartphone or tablet); the customer then authenticates a transaction by signing (in the case of signature-based transactions such as those on credit cards and some debit cards), or entering a Personal Identification Number (in the case of PIN-based transactions on some debit cards); nowadays, for some small credit-card transactions, no authentication is required.
As a result of contactless cards, there are established certification standards for NFC-enabled payments including PayPass by MasterCard, PayWave by Visa, and PayExpress by American Express, and Zip by Discover. However, these “acceptance brands”, indicating to consumers that they can use correspondingly branded contactless cards or NFC-enabled smartphones, are not widely available. Specifically, mobile research firm Aite Group forecasts that 1.3 million merchant locations will have contactless-ready terminals by year-end representing about 5% of the 24 million locations with card-swipe terminals for Visa or MasterCard; availability is better among large merchants with 25 of the top 100 national retailers have deployed or are deploying contactless terminals.
That said, NFC capabilities are now standard in check-out registers from POS providers such as Verifone and Ingenico so that, given the natural refresh cycle for POS technology, it is only a matter of time before the NFC acceptance is standard: for example, Telco data consultant Berg Insight estimates that by 2017 over 80% of cash registers in the US will be NFC-ready.
The Bank/Carrier Dispute over Access to Payment Account Information
Of course, NFC-enabled payments require not just an NFC-ready merchant terminal but also an NFC-enabled payment device capable of securely storing payment account information and, if the payment utility is to be integrated with rewards programs and transaction-tracking, information about loyalty rewards and payment activity. On a contactless card, the radio-frequency and information-storage capabilities are provided by a specialized chip but the migration of NFC to mobile payments has been delayed by the incorporation of these capabilities into smartphones as described below:
- NFC-enabled smartphones: Most high-end smartphones today (with the iPhone being an exception although probably not in the next release) incorporate NFC capabilities not so much for mobile payments as for other features including, for example: information-exchange (popularized by Samsung’s memorable “bump” commercials of last year); for device-pairing for a Bluetooth or WiFi connection; and electronic ticketing. These other functions are important to resolving the chicken-and-egg problem which would otherwise result in NFC payments with handset manufacturers unwilling to invest in the technology for payments until merchants did so, and vice versa. As it is, there are ~20 million NFC-enabled smartphones in the US representing ~15% of the total and we expect substantially all smartphones to ship with the technology by 2017.
- Secure Storage of Payment Account Information: Despite this penetration of NFC-enabled smartphones, eMarketer estimates there are only 10 million consumers who have used mobile payments. The reason is that it is not enough for a smartphone to have NFC capabilities; the developer of a digital wallet application also needs to be able to access payment account information, and this has been the subject of commercial dispute between the carriers who provide the traditional secure element for account identification (i.e. the SIM card) and bank issuers of cards who are hoping their cardholders will upload payment account information to a smartphone; the carriers want to charge the banks for their customers’ use of storage capacity on the SIM and, until recently, the parties have not been able to agree terms.
In practice, the dispute between carriers and banks over access to the SIM card for payments information has hobbled the utility of digital wallets and so contributed to the delayed roll-out of mobile payments infrastructure. The experience of Google Wallet, launched in September 2011, provides an illustrative example. Given Google’s relationship with Verizon (through Droid phones, for example) and licensing of Visa’s PayWave certification for mobile payments (so that the Google Wallet is accepted wherever merchants accept Visa’s contactless cards), as well as the involvement of Sprint, Nextel, and MasterCard as partners, Google Wallet seemed to have the ability to catalyze mobile payments adoption. However, the results have been disappointing with usage significantly behind awareness (see Exhibit 2). The reason is that Verizon, AT&T, and T-Mobile (who are partners in a competing Isis-branded digital wallet as discussed below) did not allow Google Wallet access to the SIM card for payment account information rendering it unusable on their networks.
Exhibit 2: Consumer Use of Google Wallet Lags Awareness
The National Rollout of Isis Changes the Game
After NFC-enabling the Galaxy S3 and attracting a credible wallet-developer in Google, Samsung must have been frustrated to have payments utility limited to a single network. It responded in February 2013 by announcing a partnership with Visa in which it would incorporate Visa’s PayWave certification technology directly into the Galaxy S4 allowing users to link their phones to their Visa accounts without the involvement of the carriers: in short, Samsung installed a secure element to hold payment account information directly into the phone thereby by-passing the SIM card and potentially dis-intermediating the carriers. Of course, carriers can still use their gatekeeper position with subscribers to block digital wallets but this is more brazen than arguing, as they have, that Google Wallet is not blocked but merely unable to access secure storage.
Regardless, the possibility of disintermediation may have contributed to the bring-forward of plans for a national roll-out of the Isis digital wallet which had not been expected until next year but, according to an announcement on July 31st, is now slated for later this year with the completion of trials in Austin and Salt Lake City. The inter-relationship of mobile payments and loyalty programs was striking at these trials: two-thirds of active users opted to receive offers and messages from favorite brands and fully one-third loaded a My Coke rewards card prompting Coca-Cola to comment: “working in partnership with Isis, we have been witness to the power of mobile payments and loyalty”.
Unlike Google Wallet, Isis is a partnership of bank issuers (specifically JPM and COF) as well as carriers, and it is compatible with the all the major contactless acceptance brands. As a result, we expect the rollout to establish NFC as the standard for phone-based payments, just as it is for contactless card payments, so that mobile payments do indeed follow the hockey-stick growth from $1 billion in 2013 to $10 billion in 2015 to over $50 billion in 2017 forecast by e-Marketer. In particular, we expect other carriers (such as Sprint and Nextel) to join AT&T, Verizon, and T-Mobile on the Isis standard and other card issuers (such as Bank of America and USB who until now have focused on the V.me online digital wallet offered by Visa) to join JP Morgan Chase and Capital One in embracing NFC technology.
Alternatives to NFC: Barcode and Closed-Loop
The delay to the rollout of an NFC-based mobile payments infrastructure has led to the exploration of other technologies including, in particular:
- Image-based technologies using bar-codes (such as Apple Passbook, Starbucks successful digital wallet with 3 million active users, a mobile wallet in trial by Bank of America, and the Merchant Customer Exchange (“MCX”) wallet under development by a consortium of merchants including Wal-Mart and Target. In this approach, the smartphone displays a barcode that, scanned into a barcode reader at point-of-sale, serves the same identification purpose as the magnetic stripe on a traditional credit card. A drawback to the technology is that the consumer needs to open an application and scan a barcode; the process is not more involved than digging a wallet out of your pocket and swiping a credit card, but neither is it much less involved.
- Innately closed-loop technologies (such as Square and OpenTable’s just-announced mobile platform for client restaurants) which communicate payments information through the cloud rather than directly between smartphone and merchant terminal. These technologies are innately closed-loop because they require both customer and merchant to have an account with the payment provider. As such, they do not require information exchange between a smartphone and merchant terminal provided there is an alternative means of identifying the consumer. Square achieves this by leveraging the location-based abilities of a smartphone and requiring the customer to pre-register a photo with the merchant; as a result, a Square customer can effect “hands-free” payment merely by saying his or her name which the cashier then compares with the pre-registered photo. For OpenTable, identifying and locating the customer is straightforward given the electronic reservation book (“ERB”) digitizes activity at the host-stand.
Like Square and OpenTable, PayPal uses a closed-loop technology although, as a result of its recently announced partnership with Discover, may embrace Discover’s Zip acceptance brand for NFC payments. PayPal claims 123 million “active digital wallets in the cloud” and looks to migrate these from the online to physical worlds through the Discover partnership under which PayPal branded payment devices, including a card as well as mobile version of the online digital wallet, will be accepted at over 7 million merchants after 2015 (see Exhibit 3). We do not expect PayPal will establish acceptance relationships with Visa and MasterCard, as it has with Discover, because of the potential threat to their debit businesses as discussed below; Discover does not have such a proportionately meaningful debit business.
Exhibit 3: Rollout Plans for Paypal-Branded Payments at Merchants Accepting Discover Card
Section II: The Risk of Mobile Payments to Visa and MasterCard
The Cycle of Increasing Interchange
Visa and MasterCard enjoy entrenched positions with strong consumer brands and reliable technology to manage fraud, verify funds availability at payment accounts, and reverse bad sales through charge-back procedures. Furthermore, they have a strong hand with merchants who typically cannot afford to lose a sale by not accepting a customer’s preferred form of payment; the odd exception such as Costco (which takes only American Express) proves the rule. As a result, competition in the card network business has been more about persuading banks to issue cards on the network than persuading merchants to accept them.
In practice, this means that an “interchange fee” paid by a merchant accepting a Visa or MasterCard payment card to the bank issuing the card tends to increase. The pricing dynamic leads to a perverse cycle where bank issuers use interchange proceeds to fund increasingly rich rewards to cardholders whose preference for the associated cards strengthens; the merchant is then still less able to refuse the card allowing Visa to raise interchange still further. In debit (see Exhibit 4), interchange increased until 2010 when the Durbin Amendment capped it at 21 cents plus 0.05% of the transaction value; as Ronald Congemi, a former CEO of the Star Systems one of the electronic funds transfer (“EFT”) networks competing with Visa in PIN-based debit, commented: “what we witnessed [before Durbin] was truly a perverse form of competition. They [Visa] competed on the basis of raising prices. What other industry do you know that gets away with that?”
Exhibit 4: Increasing Debit Interchange
In credit (see Exhibit 5), there is a cycle of increasing interchange particularly for premium cards where cardholders are not, in fact, accessing credit at all but rather paying the balance off every month; indeed, through funding rewards, this increasing interchange is a driver of consumer use of these cards. Merchants are understandably frustrated with the interchange mechanism through which they are funding rewards associated in the minds of customers with the brand of the bank issuing the payment card rather than with that of the merchant itself (except in the case of co-brand programs such as the Chase-Amazon card). As discussed below, digital wallets may be a catalyst for blunting this interchange cycle.
Exhibit 5: Increasing Credit Interchange
Source: GAO Report to Congressional Addressees, November 2009 (Fees assume a $40 purchase at a typical merchant)
The Merchant Response
As a result of these interchange pricing policies, Visa has created deep antagonism with merchants leading to years of litigation and a keen motivation to find alternatives. For example, in the quarter after the Durbin Amendment in April 2012 allowed merchants to choose to route PIN debit transactions over an alternative network to Visa, debit transactions on Visa’s PIN-based Interlink network fell by 50% to $48 billion; this was enough so that overall consumer debit volumes (across both signature- and PIN-based cards) grew only 1% in the twelve months through June 2012 versus 14% in the prior 12 months (see Exhibit 6).
Exhibit 6: Visa Volumes (sourced from company reports)
The transition from physical to digital wallets may give the merchants an opportunity to shift the balance of power in their favor. As discussed below, the specific risk to Visa is that merchants will use a digital wallet to interpose a proprietary card between the Visa account and the POS terminal, and then incentivize cardholders to fund the interposed card with ACH-accessed bank accounts rather than Visa accounts. The Appendix discusses Starbuck success in achieving the interposition of its store cards by a careful migration from the Duetto card (which had dual functionality as both a Visa card and a store card), to a Starbucks card without the Visa brand, to finally a Starbucks digital wallet.
Rethinking Proprietary Cards for Mobile Payments
Merchants have for a long time looked to encourage customers to use proprietary cards for payment, or simply to capture loyalty rewards, so as to increase engagement in the merchant brand and lower Visa acceptance costs. The traditional approach has a number of challenges as the cards have limited utility and quickly lead to wallet clutter. Furthermore, those cards with attached credit lines have high interest rates (because limited-purpose means low balances), giving rise to adverse selection (people who have a lower rate Visa card will prefer it), and hence credit problems.
Mobile payments can address the problems of both wallet clutter and credit availability. For example, I have a loyalty cards for Starbucks, Rite-Aid, Frannie’s Goodie Shop, and Staples: they lie in a drawer and create more frustration (because of the difficulty of finding them at the time they might be useful) than positive brand engagement. It would help if they were available in a smartphone’s digital wallet and pushed at checkout perhaps with an in-the-moment offer based on past purchases to encourage a trade-up or quick return visit. Of course, this is the functionality offered by the Starbucks digital wallet and other digital wallets that are (Apple) Passbook-enabled.
In terms of funding, it is no longer necessary for Starbucks (or E-Z Pass or Skype) to offer credit since the online accounts can be linked to a funding source, either a Visa account or an ACH-accessed bank account, which is called upon for funds as needed with cards with an e-mail notification. These various merchant accounts appear as line items on the funding-account statement and detail on the underlying purchases is available by inspecting the store account. In short, a digital wallet allows merchants to interpose their proprietary cards between Visa and checkout without any clutter or convenience to the customer.
Interposed Proprietary Cards: Transforming Rewards and Blunting the Interchange Cycle
This interposition in a digital wallet of a proprietary card between a Visa account and point-of-sale, whether using an open-loop contactless payment standard as at Starbucks or a closed-loop system as at OpenTable, is potentially transformational for both rewards programs and interchange:
- Transforming Rewards: An interposed proprietary card makes in-basket purchase information available to a merchant but not Visa (which sees only aggregate dollar amounts each time the interposed card is funded from a Visa account). Levering smartphone capabilities, merchants can use this information to offer contextually-relevant rewards based on shopping behavior, the location of the customer, and time of day. As Forrester Research has commented, customers have already come to expect this and identify it with mobile payments: “for consumers, the value of mobile payments extends beyond the convenience and simplicity of paying with their smartphone. It’s also about the value they yield, whether through integrated loyalty or personalized rewards, as today’s customer expects to be recognized as a loyal customer and rewarded for their repeat business”. A recent survey echoed this finding with nearly 70% of consumers who use a mobile wallet noting the benefit of using it for loyalty cards albeit along with proof of insurance (see Exhibit 7)
Exhibit 7: Consumer Perception of Benefits from Digital Wallets
- Reversing the Interchange Cycle: An interposed store card and associated stronger rewards program can provide a merchant with the tools to provide incentives for consumers not to use Visa as much without risking loss of business. Of course, a merchant does not reduce Visa acceptance costs if the store card is funded from a Visa account but the costs do fall to the extent it is funded by an ACH-enabled payment from a bank account. To motivate this, the merchant can offer customers who fund using, say, e-bill additional loyalty points.
Merchants will then be funding rewards that are linked in the customer’s mind to the merchant brand rather than, via the Visa interchange mechanism, rewards that are linked to the card-issuer brand. As this takes hold, the cycle of increasing card interchange described above may be blunted and could reverse: customers will have less of a preference for paying with Visa, so merchants will be more easily able to wean customers away from Visa, so interchange rates will fall to mitigate merchant push-back, so card issuers will be less able to fund rewards, so customers will have less of a preference for Visa.
Case Study of Starbucks: Migrating Away from the Visa Acceptance Brand
Like most retailers, Starbucks would prefer that – if you are going to use a payment card at all – you use their store card. The benefit to a merchant of a store card is that there are no acceptance fees to be paid to Visa and the store card, by capturing information about a customer’s purchase and storing information about offers and promotions, provides a mean to learn about a customer’s buying habits and to improve the relevance and delivery of loyalty rewards.
However, store cards have drawbacks for consumers: they are limited-purpose (typically usable at only one store) unlike the more general-purpose Visa-branded cards; they may not offer credit and, if they do, the terms are typically less attractive than those on Visa cards; and the redemption options for rewards are typically limited to items from the store-card provider. In 2003, Starbucks began to address these issues with its Duetto card which was Visa-branded (and so functioned like any general-purpose Visa card), cobranded with Chase (and so offered credit), and had a dual function as a store card. The dual functionality meant that cardholders had the benefit of a general-purpose Visa card while, to the extent the card was used at Starbucks in “store card” rather than “Visa” format, Starbucks avoided Visa’s acceptance fees and was able to use the store card to capture information about customer preference and buying patterns and to implement loyalty programs. The dual functionality was enabled by the use of two magnetic strips on the card: one for use in Visa format, and one for us in store-card format.
Starbucks preference for consumers to use the store-card format was reflected in the rewards program which offered “Duetto dollars” at a rate of 1% of spending for purchases in the Visa format but 3% of spending for purchases in the store-card format (provided customers enabled an automatic reload from the Visa account to the store-card account). In short, Starbucks created strong incentives for customers to wean off using Visa in the stores and to begin using the store card and funding the store card from a Visa account.
By February 2010, Starbucks was sufficiently confident of its store card that it discontinued Duetto and focused on a stand-alone store card which did not carry the Visa acceptance brand (see Exhibit A1). Recognizing the importance of Visa-branded credit-access and rewards to some customers, Starbucks did not pitch the stand-alone store card as an alternative to Visa but rather a complement. Specifically, the firm offered customers credits if they funded a (now stand-alone) store card with a (now stand-alone) with a Visa card. The transition was effective in that Starbucks store card transactions increased 20% in 2010 to account for $1.5 billion of purchases or ~15% of total Starbucks volume.
Exhibit A1: Duetto discontinued in 2010 and Customers Issued a Store Card without the Visa Brand
Source: Company Presentations
Of course, to the extent customers funded the Starbucks card with Visa accounts, Starbucks continued to pay Visa acceptance fees. However, for those customers not committed to their Visa accounts, Starbucks
provided other ways of funding the store card. Specifically, in October 2010, customers were able to fund their store cards using Paypal. In this case, Starbucks neither pays Visa an acceptance fee when the customer makes an in-store purchase (because the customer is not using a Visa-branded card) or when the customer funds the store card (because the customer is not funding from a Visa-settled account).
To the extent the customer is funding the Starbucks store card with a Paypal account, Visa may still receive an acceptance fee from Paypal if the Paypal account is funded from a Visa account; however, if the customer is funding the Paypal account by registering a bank account then Visa is dis-intermediated entirely.