Several industries have recently adopted a multi-cloud strategy, and banking is no stranger to this trend. While an Accenture study from 2021 found that only 8% of respondent banks’ workloads were running in the cloud, Cockroach Labs indicated that up to 88% of surveyed banks used a multi-cloud strategy in the same year.
Admittedly, it'd be hard for the top 10 fastest-growing cloud providers to have double-digit growth ranging from 14% to 67% if players in major sectors like banking were all sticking to one provider. But rather than debating how fast cloud services will catch on in banking, let’s focus on the benefits of a multi-cloud strategy for banks and how to reap them.
From Deutsche Bank and BBVA blending Oracle Cloud and Google Cloud to other financial institutions mixing in a bit of AWS, banking sector players are taking the multi-cloud route for:
Faster execution and business continuity
In many markets today, it’s increasingly crucial to be able to get in and out of a trade quickly since an instant price shift can close the window of opportunity or leave you saddled with losses. Whether it's a bank providing capital to a trader, or a bank's investment wing executing a trade, speed remains vital in this era of highly volatile markets.
But as prices fluctuate rapidly, a flurry of transactions can strain the infrastructure used to move funds between financiers, traders and any other intermediaries. The same goes for other messages, such as requests for credit line extensions, disclosures to authorities, and more. Banks can adopt a multi-cloud strategy to avoid getting caught out by overloaded servers that process requests slower or even suffer outages.
Even the top three cloud providers can't guarantee 100% uptime, especially since transaction volume surges can catch them off-guard. However, a bank with substitute infrastructure from a different provider can quickly pivot in case of a failure on one provider's end. More importantly, the bank doesn’t have to limit these switches to instances of downtime.
During normal operations, they can monitor traffic changes relative to the available computational resources and assess the impact on the time-to-finality for each transaction. This is compared to the rate of change of factors affecting trade feasibility, such as asset/commodity prices.
Once you realise that the provider is working slower, yet you'll soon need to complete trades faster, you can set a threshold transaction duration below which a switch to a less congested provider is triggered.
User-driven service modification
One of the most important efforts in the digital age of banking is the ability to quickly convert a user demand into a product or service feature. Banks often roll out new products and services, some of which may require additional steps like declaring the source of funds and the purpose of a cross-border transaction or large forex swap.
These new offers may also need extra security steps and other anti-fraud measures like multi-factor authentication per transaction, changing permission procedures when transaction requests are received from unusual or foreign locations (for users who travel often), and more.
Product owners and developers initiate some of these efforts, while others come from user requests. Either way, a bank needs to quickly build, test and deploy the modules that will handle these actions. For example, you may need to add new buttons/pages/functions to the end user's digital banking app and do the same for the portals/dashboards used by the bank's staff.
In such cases, a multi-cloud strategy may be ideal since one cloud provider will offer better solutions for some tasks in a specific process, while other tasks could be better done using another provider. For instance, your onboarding (Registration, Know-Your-Customer, etc.) may work better with a particular cloud provider because they have better storage capabilities and robust scaling.
On the other hand, this provider may not be the best regarding machine learning capabilities. So, for anti-fraud measures, you may need another provider to create and run algorithms that continuously improve at differentiating between suspicious transaction activity, accidents and actual fraud attempts.
Varying compliance demands
In this world of heightened cyber-attacks, banks should take extra precautions to ensure that different types of customer data are always safe. However, the definition of ‘safe’ varies from one region's data regulations to another's. Cloud compliance gets even trickier as more parties are involved.
For example, if your bank relies on on-premises infrastructure for certain aspects of service provision, you might need to revise your approach. Say you’re dealing with an EU citizen currently residing outside the EU, and you’re offering them a service that involves another party in a non-EU-allied region.
You'll need to be extra careful about which parts of their data are stored or processed using infrastructure outside the EU so you don't get in the GDPR's crosshairs. Occasionally, this may be as simple as opting for a provider with built-in multi-region configurations so that they can keep all your cloud operations compliant.
However, continuous compliance is a more intricate web when you're dealing with several parties in public and private institutions and charities/non-profits. One provider may not check all the compliance boxes regarding your unique agreements with different parties. This is why a multi-cloud strategy can help keep you out of trouble.
You get to evaluate different providers' compliance capabilities for various cloud features and select the most appropriate one for each transaction. Ultimately, a multi-cloud strategy helps limit the risk of litigation, fines, bans and subsequent brand reputation damage.
Cost-effective service provision
Many banks serve customers across different countries, with varying cost implications for each customer base. For example, the cost of swapping currencies may differ by region. The same can be said for interest rates on savings accounts, loan servicing fees and any charges attached to other products and services.
Inevitably, these variations also extended to customer acquisition strategies. A bank may serve similar customer types, like importers and home buyers across multiple countries, but the offers and related communication will differ.
As a result, a bank can benefit from a multi-cloud strategy to streamline customer relationship management and overall service provision. For instance, it could take a while to determine the impact of infrastructure expenditure on the ultimate cost of acquiring one customer when penetrating a new market. Accordingly, it helps to have a provider that offers a pay-as-you-go pricing model on various cloud features.
However, this issue is even more complex since several cloud providers offer different rates for storage, instances (general purposes and compute-optimised), and other details like vCPUs and RAM. They also offer distinct discounts on hourly on-demand pricing and one-year commitments. And remember, the differences in application instructions for various markets aren’t the only considerations here.
Even when looking at one specific region, you should consider the cost of managing the ever-increasing workloads and data volumes and their dispersal across different data centres relative to latency and recovery needs. For example, older backups can be situated on more distant servers, while the closest, highest-performing servers handle more recent records and real-time data ingestion.
Essentially, the most critical elements are facilitated by the more expensive infrastructure, while cheaper providers support those you need the least at any given time.
How banks can balance these multi-cloud benefits
To get the most out of a multi-cloud strategy, you need tools to help you understand how to approach your transition to a multi-cloud setup and the subsequent efforts. For starters, you should understand the new demands of deploying the same application across different clouds.
Some may be simple adjustments in utilising computational resources, while others are deeper logical changes. The goal is to limit the number of recurring manual steps involved in deployment. You also want to ensure that when serving different customer needs involving interactions between modules across multiple clouds, you orchestrate these unique processes with minimal IT help.
For example, a relationship banker should be able to rearrange a service to fulfil a customer’s request in harmony with internal and external approval procedures in just a few clicks. Another crucial consideration is the analytics side of your strategy. You’ll need to figure out all the relevant performance metrics to track, how to converge and compare them, and the journey to acting on the resultant insights.
Some analyses inform long-term strategies for multi-cloud usage, while others should be funnelled into automation modules that activate immediate changes to your cloud usage. You can achieve such efforts using platforms like Ori Global Cloud, which simplifies cluster provisioning and onboarding, application packaging, deployment and optimisation. Explore our website to see how the platform works or speak to a member of the team here >>>