[Full disclosure: This blogger worked on Google Wallet 2011-2013]
Google recently announced that its Wallet Card launched in 2013 will be discontinued this summer:
“After careful consideration, we’ve decided that we’ll no longer support the Wallet Card as of June 30. Moving forward, we want to focus on making it easier than ever to send and receive money with the Google Wallet app.”
This is the latest in a series of changes starting with the rebranding of original Google Wallet into Android Pay. It is also a good time to look back on the Wallet Card experiment in the context of the overall consumer-payments ecosystem.
Boot-strapping NFC payments
The original Google Wallet launched in 2011 was a mobile application focused on contactless payments or colloquially tap-and-pay: making purchases at bricks-and-mortar stores using the emerging Near Field Communication wireless interface on Android phones. NFC already enjoyed support from payment industry, having been anointed as the next generation payment interface combining the security of chip & PIN protocols with a modern form factor suitable for smartphones. (Despite impressive advances in manufacturing every-thinner phones, it’s still not possible to squeeze one into a credit-card slot, although LoopPay had an interesting solution to that problem.) There were already pilot projects with cards supporting NFC, typically launched without much marketing fanfare. At one point Chase and American Express shipped NFC-enabled cards. That is remarkable considering that on the whole, banks have been slow to jump on the more established contact-based chip & PIN technology. NFC involves even more moving parts. Not only must the card contain a similar chip to execute more secure payment protocols, but it requires an antenna and additional circuitry to draw power wirelessly from the field generated by a point-of-sale terminal. In engineering terms, that translates into more opportunities for a transaction to fail and leave a card-holder frustrated.
Uphill battle for NFC adoption
Payment instruments have multiple moving pieces controlled by different entities: banks issue the cards, merchants accept them as a form of payment with help from payment-processors and ultimately consumers make payments. Boot-strapping a new technology can be either an accelerating virtuous-cycle or stuck in a chicken-and-egg circularity.
- Issuers: It’s one thing for banks to be issuing NFC-enabled plastic cards on their own, quite another for those cards to be usable through Google Wallet. After all the whole point of having a card with chip is that one can not make a functioning “copy” of that card by typing in the number, expiration-date and CVC into a form. Instead the bank must cooperate with the mobile-wallet provider (in other words, Google) to provision cryptographic keys over the air into special hardware on the phone. Such integrations were far from standardized in 2011 when Wallet launched, leaving customers with only two choices: a Citibank MasterCard and a white-label prepaid card from Metabank. Not surprisingly, this was a significant limitation for consumers who were not existing Citibank customers or interested in the hassle of maintaining a prepaid card. It would have been a hard slog to scale up one issuer at a time but an even better option presented itself with the TxVia acquisition: virtual cards for relaying transactions transparently via the cloud to any existing major credit-card held by the customer. That model wasn’t without its own challenges, including unfavorable economics and fraud-risk concentration at Google. But it did solve the problem of issuer support for users.
- Merchants: Upgrading point-of-sale equipment is an upfront expense for merchants, who are reluctant to spend that money without a value proposition. For some being on the cutting edge is sufficient. When mobile wallets were new (and Google enjoyed ~3 year lead before ApplePay arrived on the scene) it was an opportunity to attract a savvy audience of early-adopters. But PR benefits only extend so far. Card networks did not help the case either: NFC transactions still incurred same costs in credit-card processing fees, even though expected fraud rates are lower for when using NFC compared to magnetic stripes which are trivially cloned.
- Users: For all the challenges of merchant adoption, there was still a decent cross-section of merchants accepting NFC payments in 2011: organic grocery-chain Whole Foods, Peet’s coffee, clothing retailer Gap, Walgreens pharmacies, even taxicabs in NYC. But merchants were far from being the only limiting factor for Google. In the US wireless carriers represented an even more formidable obstacle. With Verizon, AT&T and T-Mobile having thrown in their lot with a competing mobile payments consortium called ISIS (later renamed Softcard to avoid confusion with the terrorist group) they lobbied to block their own subscribers from installing Google Wallet on their phones.
From virtual to physical: evolution of the proxy-card
Shut out of its own customers’ devices and locked in an uneasy alliance with wireless carriers over the future of Android, Google turned to an alternative strategy to deliver a payment product with broader reach, accessible to customers who either did not have an NFC-enabled phone or could not run Google Wallet for any reason. This was going to be a regular plastic card, powered by the same virtual card technology used in NFC payments.
For all intents and purposes, it was an ordinary MasterCard that could be swiped anywhere MasterCard was accepted. It could also be used online for card-not-present purchases with CVC2 code. Under the covers, it was a prepaid-card. Consumers could only spend existing balances loaded ahead of time. There was no credit extended, no interests accruing on balances, no late fees. It did not show up on credit history or influence FICO scores.
There would still be a Google Wallet app for these users; it would show transactions and managing funding sources. But it could not be used for tap-and-pay. NFC payments— once the defining feature of this product— had been factored out from the mobile application, becoming an optional feature available to a minority of users when the stars aligned.
Prepaid vs “prepaid”
But there was one crucial difference from the NFC virtual-card: users had to fund their card ahead of time with a prepaid balance. That might seem obvious given the “prepaid” moniker, yet it was precisely a clever run-around that limitation which had made the Google Wallet NFC offering a compelling product. When users paid tapped their phone, the request to authorize that transaction was routed to Google. But before returning a thumbs up or down, Google in turn attempted to place a charge for the exact same amount on the credit-card the customer had setup in the cloud. The original payment was authorized only after this secondary transaction cleared. In effect, the consumer has just funded their virtual-card by transferring $43.98 from an existing debit/credit card, and immediately turned around to spend that balance to make a purchase which coincidentally was exactly $43.98.
Not for the plastic card: there was an explicit stored-value account to keep track of. This time around that account must “prepaid” for real, with an explicit step taken by the consumer to transfer funds from an existing bank-account or debit/credit card associated with the Google account. Not only that but using a credit card as funding source involves explicit fees to the tune of 2.9% to cover payment processing. (If the same logic applied to NFC scenario, $97 purchase at the cash-register would have been reflected as $100 charge against the original funding source.)
The economics of the plastic card necessitate this. Unlike its NFC incarnation, this product could be used at ATMs to withdraw money. If there were no fees for funding from a credit-card, it would have effectively created a loop-hole for free cash-advances: tapping into available credit on a card without generating any of the interchange fees associated with credit transactions. While having to fund in advance was a distinct disadvantage, in principle existing balance could be spent through alternative channels such as purchases from Google Store or peer-to-peer payments to other users. But none of those other use-cases involve swiping— which raises the question: what is the value proposition of a plastic card in the first place?
End of the road, or NFC reaches critical mass
In retrospect the plastic card was stuck in no man’s land. From the outset it was a temporary work-around, a bridge solution until mobile-wallets could run on every device and merchants accepted NFC consistently. That first problem was eventually solved by jettisoning the embedded secure-element chip at the heart of the controversy with wireless carriers, and falling back to a less-secure but more open alternative called host-card emulation. As for the second problem, time eventually took care of that with a helping hand from ApplePay which gave NFC a significant boost. In the end, the plastic proxy-card lived out its shelf-life, which is the eventual fate for all technologies predicated on squeezing out a few more years out of swipe transactions, including dynamic/programmable stripes and LoopPay.