I was facing an eleven hour flight back from India after an IT Industry event, had checked in and was sitting in the lounge contemplating an uncomfortable sleepless night with others I had become friendly with over the last few days. In breezed someone else from the event who happily announced he had been upgraded to first class. Took a moment before half in joke someone said ‘give you $50 for the seat’, but it was a high enough price point to show serious interest and see if there was a reaction. Then the bidding started in earnest and it was serious, to cut a long story short, it was $500 that closed the deal.
Then the airline blew it, and said that their systems couldn’t handle the seat change – again!! Seems it is some sort of one off transaction to make an upgrade and must be done before the person checks into another grade, or so they said. Actually, I reckoned it was a combination of the IT system and the airline’s attitude to ‘how things are done around here’ that showed itself in the way the staff handled the request. We had just spent the last few days at the event discussing a mixture of Business Process and the enabling architecture, and the basic issue had been the difficulties associated with change, or changing systems.
It quickly became clear that we had two groups, the majority thinking about the business process, the architecture, and the technology, around applications, and a minority who thought the approach should be entirely different using Service Oriented Architecture, or SOA. In the debate that followed what became clear was the ‘traditional’ approach is centred on ‘cost’ being the issue, at least in part because of the difficulties that adding in new applications causes to the existing systems making it expensive. The ‘SOA’ group were saying ‘time’ is the issue and that adopting a radically different approach to the problem at a process definition level allowed an SOA approach to be simpler and therefore cost wasn’t too much of an issue.
The argument was that the pace of change is continuously rising and that successful businesses know how to do the right thing at the right time, their less successful competitors are planning around steady state with periodic enforced step changes, which have to be justified on a cost basis. Using this argument the airline could have auctioned the upgrades, (there were several seats), and taken the extra money. How? Well the SOA approach would at the time I checked in asked me if I wanted them to enter a bid on my behalf for any upgrade, if so, how much I wanted to bid. Then as check in closed the winners would be called back and asked to pay their bids, then and only then would the checked in transactions be forwarded as completed to the existing legacy systems in the normal way.
The example that had come to represent this at the event discussions was an interaction between two companies, (could be a purchase, or a sale, or any number of commercial activities). Traditionally each focuses on their internal process and architecture, and it’s very difficult to interface. Using SOA a third process zone is introduced that covers the process interaction between the two parties, the passenger and airline check in as an example. Architecturally, this would be a loose coupled, long running services based transaction, that on completion would then provide a simple data entry of the optimised conclusion to the existing systems of both parties.
The conclusion of this debate was an agreement that we were trying to solve new requirements with old approaches, and, therefore defining the requirements in the wrong way. The architects present had to take the lead in helping their business centric colleagues to re think their entire approach from internal procedures to external interactions. Makes you think quite deeply, and in a different way, about the importance of the Open Group and its work on defining processes, architecture standards, and even global architects certifications. Also turns out that this conclusion is not so far away from a Harvard Business School book ‘The only sustainable edge’ by Hagel and Seely Brown in which they propose Business Strategy needs to change to embrace ‘dynamic specialisation’ of each trading transaction.