Not all business applications are equal. Some are high profile, visible and customer experience affecting. For a bank, the ATM systems, card processing platforms and Internet banking applications generally fall into this category. In my book “What Every CIO Wants“, I call these Invest category applications. Behind the scenes there are hundreds, sometimes thousands of other applications quietly working away as they have done for decades dealing with the detritus of modern banking, settlements, credit card clearing, payments processing. Much the same as it has been done for decades but now at Internet scale and Internet pace. In “What Every CIO Wants”, I call these applications Contain category applications.
The characteristics of an Invest category application is that it is well supported, capable of rapid change in functionality and capable of growth in scale and volume of use. It tends to be quite expensive to maintain an Invest category application at peak performance. The business value of these applications makes this maintenance essential.
On the other hand, Contain category applications don’t change much, don’t have much need for massive changes in functionality (after all what can you do with a batch scheduler that is innovative)? As a result Contain category applications are not as well supported, when the cost cutters come around, the quiet back office mainframe systems programmer was an obvious candidate for redundancy, outsourcing or offshoring. Control and disciplined management slipped away to the point that the Contain category application became operationally unsupportable. Despite being unsupportable the application continues to work away quietly doing a great job of keeping the payments going and supporting every aspect of the bank’s operations.
What should have happened, is that a number of insupportable Contain category applications should have been moved into Kill category and retired or replaced with something more supportable. Alternatively that last highly skilled mainframe systems programmer should not have been let go. The reality is that bonus driven cost cutters rarely listen to whinging IT techies, despite the fact that they have no understanding of the operational risks they are taking by their actions.
At RBS/NatWest the inevitable happened, the chickens came home to roost and during a simple (yes it is extremely simple) CA7 upgrade, the lack of skill and domain knowledge combined with poor operational processes including testing and rollback plans triggered a catastrophe that affects millions of customers.
This won’t be the end of IT catastrophes, there are millions of Contain category systems running our digital world, supported by old men in bath-chairs with spares from eBay. Running on a wing and a prayer, digital Russian Roulette played large.