Real data on contactless EMV in transit

[Dave Birch] At the excellent, and well-attended, Transport Ticketing 2013 event in London I saw a terrific presentation from John Hill of CUBIC reporting some early data from Transport for London (TfL’s) migration to open-loop payments. In the first instance, as you will recall, TfL began to accept contactless EMV cards on the buses and later this year they will begin to accept contactless EMV cards on the underground as well. It’s all a long way from Reg Varney and his clippies!

John was kind enough to share the detailed transaction timing data that we (i.e., Consult Hyperion) had previously seen as part of our work on the Future Ticketing Project (FTP) but were not able to share before it was made public. I’ve included John’s slide below with the kind permission of TfL. What this data shows is that the mean transaction time for bus taps is around 530ms. Of this 550ms, around 100ms goes on the terminal “polling” to determine which Oyster, ITSO (the UK’s national standard for contactless ticketing) and open-loop (EMV) products are on the customer’s card, so when you take this into account with some other overheads, it looks at if the overwhelming majority of contactless EMV transactions are taking 400ms. This is slower than Oyster (which takes about 350ms) but not a big deal. As John mentioned in his presentation, we are working with CUBIC and TfL to find ways to reduce the transaction times further so that 100ms polling is in our sights right now. Anyway, here is the key slide:

In addition to the 545ms mode peak, John’s figures clearly showed much smaller peaks at 420ms and 490ms as well as a long tail out over 900ms! This is clear evidence that different card families and form factors (ie, dual-interface vs. tags) take different times to process the transactions and that the fastest cards (at around 420ms) are twice as fast as the slowest cards. This is an unexpectedly wide variation. John politely resisted my calls from the floor for him to name and shame (!), but it boils down to this down: the fastest EMV cards are executing the transactions faster than the new Oyster card, but the majority of EMV cards are slower.

Another early learning from the TfL migration was put forward by Nick Mackie from Visa Europe in the panel (on open-loop migration in Europe) that I chaired at the event. As Nick pointed out, there’s been a problem with “collisions”. What this means is that customers who are used to waving a wallet or a purse or a pocket that has an Oyster card in it, they are now waving a wallet or a purse or a pocket that has both an Oyster and one of more contactless EMV cards in it. When the terminals see multiple cards, they (as currently configured) do nothing. There are currently far too many collisions being reported, so there’s a customer education issue here. Retailers (including transit operators) have to explain to people that if they present multiple valid cards, then nothing will happen (yet another reason for moving to mobile, where Nick pointed out that Barclaycard have “hundreds of thousands” of PayTag stickers out there and customers like them).

And here’s some more hard data from the event that might be useful: there are currently 9,000 open-loop transactions per day on the bus. This is a tiny fraction of the 6.3m transactions per day, but it is growing at 2% per annum. People seem to like it (a friend visiting from Scotland e-mailed me last weekend to say how great it was to be able just use her Barclays debit card while visiting instead of buying an Oyster card). It’s a good start.

Why does all this matter? Well, it’s because the transit experience, in many markets, is disproportionately important to the contatless trajectory. I’m might only use my contactless card twice per day on the bus, but because I use it on the bus it becomes the card in the my back pocket and I might then use it to buy coffee and lunch with contactless as well.

In any case, transit is one of the rare use cases where NFC payments add some real value to the user experience simply because it is more convenient to tap your phone as you rush through the rail gate or board the city bus rather than fumbling to find a contactless bank card or transit smart card.

[From NFC – Yes, NFC, That Speeding Train is Headed Your Way | PYMNTS.com]

Incidentally, while much of the talk yesterday was about the shift to open-loop in transit, that’s not the only model. There is indeed a completely opposite roadmap, whereby the contactless cards used for transit have such high penetration and are so well-established that it makes sense to use them instead of bank cards. Look at the Asia-Pacific region. Taiwan is following Hong Kong and Singapore in allowing the transit e-purse to be extended out into retail.

The EasyCard, a contactless smartcard system for use on the Taipei MRT system, will soon become an electronic purse that can be used to purchase small-value items… The new payment system, which will allow up to NT$10,000 (US$312.50) to be stored in the card, will be put in place a year after the Legislative Yuan passed an amendment to the Act Governing the Issuance of Electronic Stored Value Cards that paved the way for the new payment vehicle.

[From Electronic purse expected to become operational in March – Taiwan News Online]

It’s clear that, as we all know, transit is mass market, technically challenging, high profile, important and demanding. It’s at the intersection of NFC, mobile, payments and (soon) big data. I love it.

These are personal opinions and should not be misunderstood as representing the opinions of 
Consult Hyperion or any of its clients or suppliers