Why Apple Pay Suica is a success and Apple Maps is not

Inbound Apple Pay Suica user experiences are endlessly fascinating and occasionally enlightening. This tweet video captures the usual ‘whoa, that’s fast’ first time reaction.

The responses are equally interesting with a few ‘so what? we have that in (London, Moscow, China, etc.)’ which is true but it’s not the same. Almost all of them are slower, don’t have e-money functions, don’t have nationwide coverage and are not hosted natively on pay platforms like Apple Pay or Google Pay. They rely on slow buggy EMV contactless credit card transactions on transit gates instead, in short they are not transit payment platforms.

Apple Pay Suica is clearly a great service and success that has not only changed contactless payments in Japan but changed Apple as well, with Apple incorporating global FeliCa and implementing A-12 Bionic powered Express Card with power reserve technology which matches the performance of dedicated Sony FeliCa Chips on the A-Series.

What makes Apple Pay Suica a success? It is a unique layering of hardware and software that tightly integrates into a single seamless experience. At the core is the basic Suica IC card format and the transit gate system technology created by JR East and Sony in the 1980s to solve a user experience problem with magnetic commuter pass cards. Successive layers were added over time: e-money, nationwide Transit IC card interoperability, and perhaps most important of all, Mobile Suica. The Super Suica additions will further enhance the fundamental technology in 2021.

Apple Pay support arrived in October 2016, global FeliCa was added in 2017. These were 2 layers from Apple that fit perfectly and extended the entire platform with a whole new ease of use service level. The result is a service where each layer builds on and enhances the whole. This is Steve Jobs 101: work from the user experience back to the technology so that the total experience is greater than the sum of the parts.


The Apple Maps problem
Contrast this with Apple Maps. Justin O’Beirne recently published a detailed progress report of Apple’s ‘new’ (in America only) map. There was surprisingly little discussion on tech blog sites, Nick Heer was one of the few to share a few observations. O’Beirne and Heer both focus on data collection and prioritization as the core problem for Apple to fix if Apple is ever going to close the map gap with Google. I think that is a misconception that got Apple Maps in trouble in the first place.

I’ve never seen data collection as the biggest problem that Apple needs to fix. In Japan for example the data collection problem can be solved quickly by swapping out 3rd rate data suppliers with first tier JP suppliers like Zenrin who already field large data collection and verification teams. Google and Yahoo Maps Japan both use Zenrin and build on top of that solid foundation with their own data.

Integration and coordination have been, and continue to be the biggest problem. If Apple cannot do a good job integrating and coordinating different map service layers so that they build on each other, it will continue to be what it is now: a collection of loosely connected technology services that don’t work together very well and tend to pull each other down instead of up. A few examples:

  • Transit
    Apple has a very good Japanese transit data supplier Jourdan, the same one Google uses. Unfortunately the good transit data gets wasted by the limited search and sort App Maps transit UI that is completely manual, doesn’t dynamically update travel times or arrival estimates, or even provide location-based alerts when you arrive. Those kinds of integrated transit notifications on Apple Watch alone would sell a lot more devices.
  • Siri
    Siri is one the most important service layers for integrating navigation, transit and indoor maps. Unfortunately Siri is poorly connected where it should be hooked into every nook and cranny. Japanese Siri can locate the nearest station, usually, but that’s it. Siri doesn’t do transit searches or suggestions.
  • Navigation
    Turn by turn has been offered in Japan for a few years but it still basically useless without traffic information, which is still missing. Lane Guidance was only added just recently.
  • Data Duplication
    This happens all the time as Apple fails at coordinating and verifying data sets from different JP suppliers.

And so on. I included data duplication as it illustrates my basic point that no matter how good the basic data collection is, it’s worthless without a robust integration and coordination process. A smart team of human editors with deep local knowledge understand how services should connect, what works and how it should work. A truly  great team also knows how to focus and do more, much more, with less. This is impossible to achieve with the current one size fits all mentality.

Apple Maps Japan is a classic ‘the total is less than the sum of its parts‘ product. To be sure there are some good parts, but in Japan they don’t add up. The different layers stay separate and never integrate into a seamless whole like Apple Pay Suica does. It’s great that Apple is making process with its map reboot effort in America but the real test will be how well they integrate it all. A laser focus and smart integration is the only way Apple can close the map gap with Google.

Advertisements

The Contactless Payment Turf Wars: why Oyster is missing from mobile

Open Loop EMV
It is very strange that the TfL Oyster card, which completely transformed London area transit still isn’t hosted natively on Apple Pay or Google Pay. Other MIFARE based cards are hosted on both digital wallet platforms and TfL has an Oyster app for account management and online recharge (top-ups). From a technical standpoint there doesn’t seem to be any particular problem preventing them. Perhaps it is a political thing.

TfL decided in 2011 to put their resources into the emerging EMV contactless standard. The reason was simple:

The current Oyster system, though very popular, is expensive and complex to administer. Contactless bank cards use existing technology, responsibility for issuing cards would lie with the banks rather than TfL, and the operating costs should be lower.

That is politician think, not business think: everything is a budget problem, not a business opportunity that needs investment, reduce costs by letting someone else pick up the tab but let them take their cut first. I wonder if TfL publishes how much they pay out in transaction processing fees to banks and Cubic? Perhaps not. Meanwhile budget pressures are not letting up as Londonist notes:

In 2017 there was a push to nudge people away from their Oyster cards and towards contactless. One announcement rang out all over London’s tube stations: Why not use your contactless bank card today? Never top up again, and it’s the same fare as Oyster.

The die was cast in 2014 and probably won’t change. Instead of putting resources into hosting Oyster on Apple Pay or Google Pay, TfL and Cubic already have a mobile solution which is ‘open loop’ ticketing with EMV contactless bank cards. Open loop does not address the finer issues of different fare schedules (children, seniors, etc.), commuter passes, season tickets, nationwide transit interoperability, regional promotion, nor does it offer the business advantages of a transit payment platform, Express Cards with power reserve or any kind of future vision. That’s the end of the open loop story because EMV contactless is a very dumb smart card.

It’s a shame really because TfL loves to say they generate the most transactions in all of Europe. To me that’s a gold mine to build an empire, budget problems solved. Unfortunately TfL gives that gold mine away to banks and Cubic.

You can see the same thinking with Oyster’s Australian cousin, the Opal card system, built and managed by Cubic, just like Oyster. Opal is also going the ‘open loop’ route instead of transit cards on mobile.

Open Loop QR

Hong Kong’s Octopus (FeliCa) and Singapore’s EZ-Link (Ex-FeliCa now CEPAS) are going open loop but in different ways. EZ-Link has been testing EMV contactless for over 2 years now, users report a less than smooth experience. Kaohsiung Rapid Transit in Taiwan which uses MIFARE based iPASS and EasyCard is also considering EMV contactless open loop while the recently opened Taoyuan Airport MRT offers QR Codes and a cute YouTube video.

Hong Kong going the QR Code route shows how badly AliPay wants in on Hong Kong transit, and MTR Corporation in on China transit, bad enough that Hong Kong will sacrifice a great transit payment platform for AliPay, another gold mine giveaway. Judging by the AliPay branding and retrofitted QR Code readers on Hangzhou Metro gates in the pictures above, what AliPay wants, AliPay gets, but the fast FeliCa based Octopus smart card stands in the way. Instead of improving Octopus or extending mobile Smart Octopus, it looks like Hong Kong will invest in very slow and very dumb QR. The Hong Kong Economic Journal had this to say about the development:

MTR has set its sights on a major revamp of its fare collection system, accepting new electronic payments methods rather than just single journey tickets and Octopus Cards. From the passengers’ perspective, it means there will be no need to have an Octopus card on hand for a journey on local trains, if MTR’s new fare collection system supports all the mainstream contactless payment methods such as Visa payWave and MasterCard PayPass, or mobile payment means like Apple Pay, Samsung Pay and Google Pay.

Japan in the middle
TfL/Oyster and Transport for NSW/Opal, Octopus, EZ-Link are government held transit authorities, not private independent companies. Publicly run transit authorities are subject to politics and special interests like any government agency, this sometimes leads to poor decisions and short-term thinking.

Japan was fortunate that major transit players like JR East, are private companies with strong technology partners, like Sony and NTT Docomo. Out of this fortunate set of circumstances Suica was created and finally reached  the market in 2001 (a fascinating engineering story). Suica became the Japan IC Transit card template which evolved into the ubiquitous Japan Transit IC Mutual Use Association project for transit and e-money use. Mobile Suica was introduced in 2006 by NTT Docomo and now resides on the Apple Pay and Google Pay platforms.

The ubiquity and scale of interoperable transit IC cards sets Japan apart from all other countries. China copied the Japanese model for China T-Union but the cards cannot be used as e-money and have been upstaged by AliPay and WeChat Pay which, surprise, can be used for e-money and transit.

Japan occupies a very unusual middle ground between EMV contactless from the West and QR Codes from China, neither of which play well together. The scale of Suica provides the breathing space for Japan to pick and chose what works best for, and enhances their transit payment platform. The result is an incredibly rich and varied contactless payments market anchored around Suica and similar FeliCa prepaid cards.

Future trends
For every marketing report that predicts QR Code payments growing into a 70 billion USD sized market by 2023, someone else calls it nonsense because Suica is becoming the card for everything. In many ways Suica already is. MITI said it is investigating using QR Codes for small rural transit systems that cannot afford IC card systems. This loops back to TfL complaint that IC cards are expensive to issue and manage.

Low-cost QR Codes certainly make sense for lightly used rural transit operations but they have a fatal weakness: they don’t have plastic card versions that work anywhere and seniors prefer the simplicity of plastic, QR Codes require a high cost network connected smart device, an app and are strictly one way read with no offline processing.

JR East and Sony have announced that they will solve cost problems for rural transit and much more in early 2021 with Super Suica.


Update: Open Loop QR Code Security Risks
One issue that was in the back of my mind while writing this post was the privacy and security implications of letting AliPay inside with direct transactions on transit gates. Japanese customers are very sensitive about where and how transaction records are held and used but I have yet to see any security discussion in connection with Hong Kong MTR opening up transit gates to AliPay and WeChat Pay. QR Code transactions are very different from offline FeliCa Octopus transactions. Where and how does the QR Code transaction data from Hong Kong MTR transit gates get stored, does the Chinese government has access to it to gather intelligence from transaction and location records?

If there is one thing we do know about Chinese companies is that they do what they want when nobody is looking. Witness China Telecom spoofing the BGP protocol to poison internet routes and suck up massive amounts of American and Canadian internet traffic for intelligence analysis. If I was living in Hong Kong I would be concerned about the privacy implications of MTR going open loop with QR codes.

iOS 12 Apple Pay Wallet pulled a MIFARE and nobody noticed

The Apple Wallet Ponta card launch at LAWSON presents another dilemma: just what exactly is Apple using for iOS 12/watchOS 5 Apple Wallet Passes and Student ID cards? Student ID cards and Apple Wallet Ponta have the same device eligibility specs: iOS 12/watch OS 5 running on iPhone 6 and later/Apple Watch Series 1 and later.

You might assume that Apple Wallet Ponta is FeliCa but the eligible device list tells a different story. You might also assume that everything in Japan is FeliCa but this is also not the case. Doutor Coffee shops sell a handy little Doutor pre-paid card that is MIFARE and it works flawlessly side by side with FeliCa flavored Apple Pay Suica on the same NFC reader.

Altogether we have an interesting spec list for Student ID and Mobile Ponta cards.

  • The same eligible device specs that only support NFC A-B across all devices
  • Stored value
  • iOS 12 PassKIT NFC Certificates
  • Express Card capable
  • Local offline transactions

I’m calling it (again): the only technology that fits this profile (for Student ID cards but not Ponta) is MIFARE iOS 12 PassKIT Wallet passes are simply MIFARE. Only Apple could pull this kind of ‘under the hood thing’ off in iOS 12 without anybody suspecting and it neatly puts all the major NFC technology pieces on Apple Pay: EMV, FeliCa, MIFARE and China Transit.

Blackboard supplies the technology and backend services for Student ID cards on iOS 12. I contacted Blackboard PR to confirm if the card technology was FeliCa or MIFARE but did not receive an answer. However I did run across an interesting Blackboard press release from 2015 Blackboard and NXP Semiconductors Collaborate to Strengthen Campus Card Technology:

Blackboards’ push to adopt NFC in addition to their existing MIFARE-based solutions, back in 2012 showed incredible insight into the potential of this technology. The security, convenience and flexibility that NXPs NFC and MIFARE solutions bring truly reflect the student lifestyle. Now access to campus services can be simply enabled via a smart watch or smart phone.

Based on this and the fact that it came 2 years after a FeliCa demo of Blackboard Student ID cards with a rumored migration from FeliCa to MIFARE, plus the eligible device specs, my conclusion is that Student ID cards on iOS 12 are MIFARE HCE (Host Card Emulation) which is NFC-A.

Apple Wallet Ponta cards on iOS 12 are VAS protocol contactless passes outlined at WWDC18 , WWDC16, and in the Contactless Passes section of the iOS Security Guide:

Wallet supports the value added service (VAS) protocol for transmitting data from supported passes to compatible NFC terminals. The VAS protocol can be implemented on contactless terminals and uses NFC to communicate with supported Apple devices.

This is also NFC-A. Contactless passes have been around for a while on iOS but adoption has been slow. With iOS 12 PASSKit, Apple is encouraging developers to migrate from QR Codes to NFC contactless passes and hopefully lowering the NFC Certificate requirement bar a little. Part of the reason for the slow uptake is poor NFC reader support. LAWSON has a new POS system built around Panasonic JT-R600CR readers which are Apple Pay savvy and Apple Wallet Ponta cards only work correctly when you tell the LAWSON cashier to use “Apple Pay”.

Update: A highly trusted NFC engineering source contacted me that I got it partly wrong. The correction edit above explains that Wallet Ponta cards are Apple’s implementation of the VAS protocol and not MIFARE. Student ID cards are almost certainly MIFARE/PASSKit NFC Certificate Host Card Emulation (HCE), Apple has not publicly announced MIFARE support but it is the only technology compatible with Blackboard IC card formats that could power the express card features of iOS 12 student ID cards across all eligible devices. Research and confirmation efforts are ongoing.

PAYGATE Station Android based smart payment reader

It’s rather unusual that a company would hold a press conference for a payment terminal but Daiwa House Group subsidiary Royal Gate Inc did just that last week for their new PAYGATE Station smart reader. Japanese IT journalist Satoshi Tanaka covered the rollout for IT Media.

The ‘thin client’ mobile Android based smart reader handles every payment protocol there is.

Contactless: EMV, FeliCa, MIFARE
Hardware: EMV IC, Magnetic stripe
QR Codes: D-Harai, Pay Pay, Line Pay, Ali Pay, WeChat Pay, Origami Pay, Rakuten Pay
Reward Cards: D-Point, Ponta, Rakuten

GMO Payment Gateway provides the payments processing backend for Royal Gate and Daiwa House Financial handles the business client support side. The product video shows the business target: small cafes, restaurants, boutiques. With the ever-widening rollout of EMV contactless and QR Code payments, Japan is turning into a very interesting and unique intersection point of Western EMV payments, home-grown FeliCa and China market driven QR Code payment schemes.

I don’t think having home grown QR Code payment systems around is a bad thing even though most Japanese don’t use them. The value is that it gives Japanese banks and companies a bargaining tool to use with Apple, Google, Amazon, Alibaba Group and WeChat. Never get in a sword fight without a sword.

Traveling abroad with Apple Pay NFC switching credit cards

Japanese tech journalist Satoru Nakayama who posted his experience of riding the Shanghai subway with an Apple Pay Express Transit China card loaded on his iPhone earlier, continues his Apple Pay NFC switching global travel adventures with trips to Taipei and London. His Apple Pay Rakuten Mastercard which uses the FeliCa QUICPay network in Japan, worked perfectly with Mastercard Contactless readers overseas.

He noted that Mastercard Contactless is making rapid progress in Taiwan, encountering none of the Mastercard Apple Pay refusals he experienced earlier this year. London was also easy and he rode the London Underground without an Oyster card (MIFARE), using Apple Pay instead (EMV contactless). His picture highlights one shortcoming of using ‘Open Loop’ EMV contactless credit card payment for transit: no Express Cards.

Another problem Nakayama san encountered is that many London stores only accept contactless payment for purchases over 5 GBP or restaurants that only accept contactless payment for 2 people or more. This is a problem that Japan’s Osaifu-Keitai solved long ago: the “e-money” contactless payments system that Sony and NTT Docomo pioneered in 2004 covered both credit cards and prepaid cards. It also eliminated signatures, PIN codes and ‘no contactless payments on sales less than XX’ nonsense. Sony and NTT Docomo knew that people would not use contactless payments if they were not accepted everywhere without conditions, just like cash.