Cloud computing will save banks billions. Here's how


Paul Taylor, Chief Executive Officer, Thought Machine

The governor of the Bank of England estimates banks can cut costs 30 to 50 per cent by moving to the cloud. A huge sum. It is enough to determine which banks survive and prosper, and which go to the wall.

Yet I believe this to be an underestimate. With the right approach banks can save more than 90 per cent.

A bold claim? Maybe. But one rooted in fact. In this blog I am going to explain the logic behind the number.

The scale of the problem

Let’s start with something we can agree on – banks are struggling with technology. The programming languages, the operating systems, the hardware, the way they test software, the way they run their teams: are all in need of modernisation.

Banks will confirm this. They know they are marooned on obsolete technology. Around 30 years ago the banking world diverged from mainstream tech, and went down an evolutionary dead end. The applications they use now are slow, inflexible, and written in languages computer graduates don’t understand.

To make matters worse, every round of M&A multiplied the complexity of back-ends as incompatible systems collided. Banks talk about “spaghetti systems” to describe the tangled mess of hundreds of applications patched together. Banks spend a fortune just to keep these systems alive.

For consumers, the results are dire. Banks simply can’t serve their customers. On average there is an unplanned outage every day. Mortgage interest payments are billed incorrectly, ATMs malfunction, and accounts are inaccessible. A recent UK Treasury Committee concluded: “The number of IT failures that have occurred in the financial services sector…and the harm caused to consumers is unacceptable.”

What really highlights the failure of banks is the comparison with the tech world. For example, Google, Netflix, and Amazon run their platforms with barely seconds of downtime a year. Their operational efficiency is orders of magnitude better than banks.

So the performance of banks looks worse and worse over time. The gulf just keeps growing.

Alas, banks are stuck. They can’t modernise their systems in any meaningful way.

Thus, banks are starting afresh on new platforms – either with a greenfield bank or a complete migration. And both solutions live in the cloud.

Cloud is only part of the equation

Now let’s talk about that 90 per cent cost saving. There’s an assumption that moving hosting to the cloud will lead to substantial savings – it’s an industry cliché. The truth is cloud hosting, be it Amazon Web Services, Microsoft Azure, Google Cloud Platform, or IBM Cloud, is only cheaper than on premise hosting - notably in staffing costs - but not a game changer in itself.

To achieve huge cost savings two more things need to happen.

1. Re-engineer software on Cloud Native principles.

Moving to the cloud is a chance to begin afresh. All cutting-edge tech companies run in the cloud on Cloud Native software.

Cloud Native is a precise term. It describes software conceived and built to take advantage of the unique properties of the cloud. I’m talking about containerisation with Kubernetes orchestration, to enable near infinite scaling and rock-solid stability. Microservices – in which applications are broken into autonomous chunks, which communicate via APIs.

Thought Machine is a Cloud Native company. Our core banking platform Vault Core is built with these cutting-edge technologies. Performance, security, and stability are all radically improved.

Running Cloud Native software is significantly more efficient, both in terms of performance and staffing costs. Maintenance duties are cut to near zero: a core principle of Cloud Native is to be “self-healing”. Malfunctioning instances are spun-down automatically and replaced by new instances. Many of the tasks traditionally associated with running an IT infrastructure are redundant. IT staff are left to focus on innovation.

The cost of integrating with third-party services is reduced with Cloud Native software. APIs make connectivity easy. Adding a new payment service, or artificial intelligence engine, can be done in weeks - impossible with traditional software architectures.

But even with optimum Cloud Native design we are still only at 20 per cent of the potential cost savings.

2. Adopt a Technology Company approach

The bulk of the gains from migrating to the cloud arrive by adopting what I call a Technology Company approach. This means operating the bank with the same cutting-edge concepts found at Google, Amazon, Netflix, and other elite tech companies.

These are ideas I learned first-hand during my time at Google with my co-founders Will Montgomery, Peter Ebden and Fabian Siddiqi. The Technology Company approach is a combination of culture and practices. And they are radically different to the way banks work today.

Adopt Technology Company principles and the savings really can hit 90 per cent.

Life as a Technology Company

A key, defining principle of Technology Company thinking is automation. Anything and everything is automated. This allows for incredible scaling and efficiency. When I worked on the Android team at Google, we had one team member per million users. WhatsApp is reported to have one developer per 10 million users.

To act like a Technology Company you need a purity of purpose, pursuing automation ruthlessly. All manual processes must be eliminated. Banks, unfortunately, are very keen to throw people at problems when they should be throwing machines at problems.

Automation should be at the heart of how code is written and tested. While all companies, including banks, use code repositories (“repo”) like git, they often have one repo per project or component. This allows teams to work quickly and independently at the start, but it stores up problems for the future. Google, by contrast, uses a single repo for all its main code, and Thought Machine uses an identical approach. This is called “monorepo”. Monorepo is combined with what’s known at Continuous Integration, Continuous Development (CI/CD). This means that when a developer checks in a bit of code, it automatically integrates with everything else that’s there. It deploys straight away into a pre-production environment. Testing is done automatically.

I want to stress how important this last part is. Automatic testing can be done hundreds of times a day – impossible with manual testing. For developers, this is revolutionary. It means bugs can be identified immediately. Compare this to the way banks currently work – launching Big Bang upgrades every six months, so if something breaks the developers won’t have a clue where the problem lies.

With full automation it is possible to deploy code updates hundreds of times a day. Innovation happens via a blizzard of incremental improvements. Amazon, for example, revealed it deploys new code every 11.6 seconds. Etsy deploys code more than 50 times a day – an approach requiring 13,700 automated tests a day.

This is how Technology Companies work. It is how banks need to work.

Consider the advantages for deploying CI/CD, in a fully automated environment.

  • Applications will evolve over time by incremental improvement. No longer will a bank risk being marooned on out-of-date software, unable to upgrade because the gap has grown too large.
  • Innovation is made easy. Product designers can make valuable changes to the customer experience on the fly. A/B test. Launch new products in weeks, not months.
  • Improved stability. Introducing small, gentle, frequent changes is far more reliable than Big Bang upgrades once every six months. We all know the bank’s track record on these mega-releases – they have the potential to ruin the reputation of a bank.
  • Staffing costs are reduced to a fraction. In a fully automated bank all routine processes are delegated to machines.

Google and Netflix offer proof

So I think the Bank of England is being conservative with a cost saving estimate of just 30 to 50 per cent. When we compare the efficiency of the best Technology Companies we can see banks stand to gain 90 per cent by adopting the same approach.

We at Thought Machine are helping our clients, including Lloyds Banking Group, SEB, Atom Bank, and Standard Chartered, enter this new world. They are transforming into Technology Companies, with the same philosophy as Google and Amazon.

The gains are astronomical. We hope other banks embrace the future, and join us.

<< Previous blog
Next blog >>
Core modernisation and ecosystem collaboration with Accenture
Read this blog
Empowering our clients to accelerate their modernisation
Read this blog
Money20/20 Banking Infrastructure Summit: What to expect with a core modernization program
Read this blog
To be ISO 20022 ‘ready’ is not good enough
Read this blog
Building a winning team with a strong culture
Read this blog
Domain-driven design and the future of payments
Read this blog
How does Vault Core compare to closed-box systems, and what does this mean for product development?
Read this blog
Introducing our Enablement Portal – a complete resource for support, knowledge and training on the Vault platform
Read this blog
The Integration Library: a growing collection of solutions with best-in-class technology vendors
Read this blog
Building a bank on top of kubernetes
Read this blog
From speech technology to banking
Read this blog
Strategic partnership with Lloyds Banking Group
Read this blog
Cloud computing will save banks billions. Here's how
Read this blog
A demand for COBOL expertise underlines the fragility of critical infrastructure
Read this blog
Why are cloud systems so much more reliable?
Read this blog
Life may have slowed down but innovation doesn’t!
Read this blog
Building a core banking system in a distributed environment
Read this blog
Strengthening our commitment to cloud native design
Read this blog
Cloud Native - what does it mean? An interview with CNCF's Cheryl Hung
Read this blog
Shaping the future of banking IT: We’ve joined the Banking Industry Architecture Network (BIAN)
Read this blog
Why microservices are the future of banking
Read this blog
GFT and Thought Machine forge strategic partnership to accelerate global banking transformation programmes
Read this blog
Q&A with Nick Wilde, MD of Thought Machine Asia-Pacific
Read this blog
How Thought Machine can unlock the cloud for banks with Red Hat OpenShift
Read this blog
Round table: Meeting the challenge of a digital future
Read this blog
How to go full cloud native with CockroachDB
Read this blog
Meet our chair Andy Maguire
Read this blog
Core banking transformed: accelerating migrations with cloud-native cores
Read this blog
Let business justify your investments into digital-native core banking systems
Read this blog
How Thought Machine exceeds global industry standards in compliance and security
Read this blog
Thought Machine redefining banking with Standard Chartered
Read this blog
Sign up to our newsletter
Thank you! You will now receive some incredible content in your inbox!
Oops! Something went wrong while submitting the form.
For information about how we use your data please read our privacy policy.