8 Sept 2013 (sorry for typos)
Read in the press today about ISIS’ new SmartTap protocol. I may be getting to an age where I feel everything is on replay.. as Yogi Berra said “Its Déjà vu all over again”. How many ways are there to integrate to the POS? My simple pic below outlines 10-12. Integration is NOT a technical problem.. it is a BUSINESS VALUE PROBLEM.
Given we are all biased by our life experiences, thought I would share mine. I was very fortunate to work for Oracle from 98-03 and had a very, very good team… one of the areas I led was the solution architecture practice. In order for Oracle to sell software we had to define how it worked with everything else. We may have had the best Advanced Planning and Scheduling (APS) or CRM in the business, but how does it work with SAP SCM, or Peoplesoft Financials? Sometimes the challenges were all Oracle, I can still remember meeting the new Motorolla CIO in 2002, where he said “we have 124 ERP instances of different versions.. and they are all Oracle.. now it is our fault for buying the same product 126 times… but it is your fault for selling it to us… We can’t create a single integrated view of our company.. where do we start?”.
Perhaps my biggest “ah hah” moment came from working within HighTech Manufacturing (ex Cisco, Sony, TSMC, Samsung, …). The need for coordination in the supply chain was EXTREME (to avoid the bull whip effect). RosettaNet was born… a set of XML messages that could be exchanged between participants communicating supply chain information like Work in Process, Shipment notifications, purchases, demand plans, … There was nothing wrong with this specification… technically it was rock solid (much like NFC), but the adoption was terrible. Two reasons
1) Business. I was working with a supplier who was manually keying in WIP responses that they believed would placate Cisco (the Channel Master). When asked why he was doing this manually vs letting Oracle’s system do it he responded “if Cisco sees me falling behind in WIP they will shift demand immediately to another contract manufacturer…. I cannot risk having a system send this message to my most important customer”. In essence there were no business incentives for participation. The “mandates” for participation were end run.. Thus demonstrating that technical integration is perhaps the smallest problem which exists in networking businesses.
2) Technical. To use our RosettaNet Connector you had to be on the latest version of our applications and Infrastructure (DB/Middleware). Less than 10% of Oracle’s HT customers were on a current release, as most heavily customized. Integrating a “connector” into the OS/Application was MUCH different than having a customer turn it on and use it. This dynamic seems to be forgotten by every enterprise software partnership/alliance.
With respect to ISIS, NFC and SmartTap. Everyone is learning the same lessons over and over again (business and technical).. For example, Micros is building “connectors” for ISIS, PayPal, Google, …etc. Verifone is also building “applications” for the same. There is a lost truism: Building technical capability is MUCH different than using it (particularly if merchants must upgrade software and sign a new contract.. all for 2% of their customers that may have a new ISIS NFC phone). ISIS will also learn that partners like Micros do not sell the adapters they build.. nor do merchant have the latest version of their software. Merchant USE of an ISIS SmartTap adapter is MUCH different than building it.. Same holds true for Google, Paypal, … etc.
Take a look at the diagram above.. I didn’t even bother listing “integrations” from the payment terminal, yet Verifone established a Verix architecture to enable a segmented area where non payment applications could run (within a payment terminal). Google built an adapter here in Wallet 1.0. Someone should ask Verifone how many applications run in this environment.. and how many merchants integrate back end non-payment data to a payment terminal (ex line items, loyalty, …). Answer is very close to 0…. the idea that a merchant would use a payment terminal as a consumer integration point is just ludicrous. The very existence of a specialized payment terminal is due to the need for “specialization”.
Cloud is where Payment belongs
As I stated in Battle of the Cloud – Part 3, if everything is connected, why on earth would I want to store anything in the phone? Everything should be in the cloud.. and all that is needed it a form of authentication to pull in everything I know about the customer. For retailers why would you want a teleco to manage your loyalty program for you when a Starbucks/Square/Apple Passbook paradigms shows QR code options? For that matter why would you do anything that Apple is not supporting? (the best most affluent customers).
Riding existing rails seems like an obvious approach to circumvent some of these integration challenges. But current network rules and participation are based upon existing data flows, current rails are poorly equipped to handle new business models. For example I asked a top 3 Retailer if they would ever share transaction data with Google. They said if it was good for their customer and good for them then perhaps…. But this data would never ever flow though visa, the banks or any other payment network… as they have proven to be very bad partners in commerce.
Thus my view on why the problem is NOT technical.. just ask one simple question.. ISIS what does this do that Apple Passbook or Google wallet can’t do already? Isis should have asked micros… How many retailers use these kinds of integrations (of the type I’m building). Micros answer…. We have no idea. .. We don’t get paid when they are turned on…
On the technical front I will make a predictions.. not only will ISIS realize retailers don’t want to turn this on.. Customers will reject the solution as SmartTap requires 2 serial NFC transactions, first payment, then for “other” .. meaning a tap and “Hold” for up to 3 seconds. If the customer just wiggles their phone around during that 2-3 second time the second transaction is lost. This is what happened to Google wallet 1.0 2 yrs ago..