The Digital Transformation in Financial Services, a Tale of Two Banks.

Recently I ended a 3-year contract at a large American bank. I knew the bank was behind in its 'digital transformation' but I just didn't know how far behind it was.

I knew they were behind because we (business IT) spent anywhere from 20-50% of our time in dealing with infrastructure issues such as login registration, login troubleshooting (logins often didn't work), software installations, upgrades, CI-CD setup and use, security issues and myriad other things not directly related to the business systems we were developing and supporting.

I got a clue this week about how far we were. I was interviewed for a position of a solution architect for a large European multinational financial services company (equipment leasing and financing) owned by an European commercial bank.

I never heard of the company before. So I did my research, I parsed the job description trying to guess what kind of business and technical environment they could have and hence what were the relevant skills I would want to showcase.

And I made a fatal assumption, I thought that since the company was little known and worked in the background it couldn't possibly be very sophisticated at least in terms of its IT technology. Besides I already worked in a similar business environment when I worked for a leasing division of HP some 20 years ago. In those days HP Finance had one large global leasing system (where payments and assets were tracked), custom developed in-house running on HP 3000, one e-commerce system (which was obsolete within the first year of production that's how fast e-comm was moving then), there was also a credit decision system which was fairly good after we completely re-factored the original COTS system. Based on these assumptions and my previous experience I expected a small IT shop focused on small Canadian market with a few small systems that perhaps might need to be integrated or upgraded.

Well, I was in for a shock. It turned out that what looked like a little regional outfit turned out to be a sophisticated banking operation working globally, using EA (Enterprise Architecture) as its driving methodology to move all its capabilities on-line across the business landscape (Azure, AWS, GCP) using all the latest technologies, methods, strategies, architectures to achieve all the cost savings and functional advantages available. The little outfit had 70 apps or components running on cloud and adding to the count daily. All running for peanuts in terms of costs. And they employed a handful of IT resources (alas people) who managed all this complexity on a daily basis. This is the impression I got from a short one-hour conversation with the senior architect so I am sure it wasn't as rosy as it sounded but I also don't doubt they were achieving major shift from private hardware-based ecology to a continuous digital landscape.

After I recovered from the initial shell-shock I thought of the major achievements at the bank while I was there in terms of their "digital transformation". Well, they did upgrade their ancient RHEL6 Linux hosts to RHEL8 virtual servers which took about 2 years. On one level this can be seen as step forward in that a more modern environment is available for the business systems so software can be upgraded (we used PhantomJS but couldn't use the latest version simply because it didn't run on RHEL6), more (virtual) servers can be provisioned quickly. However in terms of 'digital transformation' it's actually a step back because now-days no organization should be spending any money on hardware unless they are doing something so esoteric that it's simply not feasible to do in the could.

Cloud is today's hardware or more precisely a vast warehouse of hardware where you come in and select exactly what you need for the amount of time and work you need it. Money spent on actual physical hardware is money wasted, at least most of it. One tenth of that money could have bought the same horsepower on the cloud and the remaining 90% could have gone to create additional significant business functionality or more and better testing or any number of other 'soft' things. Instead the business is now tied for years to their hardware until it's rendered either obsolete functionally or the cost had been fully written off.

But it's not just a hardware issue, this has a cascading effects on the overall business environment. When you buy your own hardware you need a support team which brings additional significant costs that has to be subtracted from the overall IT budget. And here is the real killer of new hardware, hardware use is difficult, slow and expensive; users have to be setup, permissions have to be granted, this has to be maintained, supported. Software needs to be installed, upgraded, migrated, patched etc. etc.

Simply put, a 'digital transformation' that includes significant hardware purchases is in software development parlance an 'anti-pattern'. In doing so you effectively give all your competitors a large head start that you may never overcome. While you will be spending money to support your expensive hardware your competitors will be building business services for themselves and their customers, year after year. The functional accumulation of additional business capability may be so large that you will never be able to overcome it and will be permanently stuck in playing catch-up, all other things being equal.