As one of the first 10,000 users of Netflix, I’ve been following their progress closely for more than a decade. Netflix was early to the streaming game and began building their data centers for not only the faster distribution of their content, rather than using the postal service but also to manufacture their content, rather than DVD factories. This makes financial services and Netflix very alike, as the two use information technology (IT) for distributing and manufacturing their products.
Although what we now think of as Cloud started with the creation of Amazon Web Services (AWS) in 2006, for many of us it first began to seem important when Netflix replaced all their data centers with AWS and effectively outsourced their value chain. This prompted the question, why? According to Adrian Cockcroft, then chief architect at Netflix, it wasn’t about cost, although using AWS did save them money, it wasn’t about outsourcing although they no longer needed to run their own data centers, it was about not waiting. It was about unfettered scaling; they could provision a cluster of 500 virtual machines in five minutes, freeing them from the constraints of their business model.
Taking the waiting out of wanting
Financial services, and particularly banks, are going through a similar change. They have replaced their data centers with the cloud and use software-as-a-service (SaaS) to manufacture and distribute the products and services they provide to their customers whether in retail, corporate, or wealth banking. As a result, just like Netflix, the bank can take the waiting out of wanting, as the old credit card adverts used to say. This is becoming important to banks, as services become near real-time, in effect becoming ‘on-demand’ services. Faster Payments in the UK, FedNow in the US, Unified Payments Interface in India, and Instant SEPA are examples of this for real-time payments. But payments are not the only banking service moving to near real-time. Buy Now Pay Later (BNPL) offerings are making retail loans real-time. Increasingly, the banks’ customers are expecting a near real-time on-demand service 24 hours a day.
This is becoming important to banks, as services become near real-time, in effect becoming ‘on-demand’ services.
On-demand resource provisioning offers another attraction to banks. Currently core banking services are sized around the overnight interest capitalization job that banks typically run once a month. A typical customer needs a large multiple core XEON machine, running at a high utilization to complete this job. The rest of the month it runs at around 8% utilization. Moving the job to on-demand provisioning can do two things for the bank: it can significantly cut the infrastructure cost by a factor of twelve and can reduce the time for doing the job to just a few minutes.
Different approaches for different banks
Moving to SaaS banking software might take different approaches for different size banks. Large banks typically have multiple application IT departments and a central IT department that runs the data centers and provides guidance to the application teams on what platforms to use. Replacing these data centers with cloud infrastructure allows the central IT team to provide the platforms for the application teams to use, providing more flexible and consolidated operations. Very large banks will build these platforms, other large banks will rent them. Small banks will start to find it very compelling to move from renting platforms and building their applications to renting the applications too. This is software as a service.
In conclusion, there are several benefits of software-as-a-service for banks looking to keep up with customer demands. As the industry moves towards near-real-time and on-demand services, the acceleration of technology will continue to play a major role in shaping finance. Eventually, all banks are expected to follow Netflix and adopt shared infrastructure.