Historically, the high street branch was the main channel of distribution for banks. Over time this was augmented by ATMs, telebanking and eventually mobile banking.
But as technology has evolved, the pace of change has accelerated. Customer adoption of digital channels, expectations of richer online experiences, external pressure from big tech and challengers have forced banks to innovate. Business and operating models are now firmly ‘digital first’, leveraging new distribution channels and routes to market to maintain key drivers like Net Promoter Score (NPS) – and to see off competitors.
Pressure on margins
Low interest rates and escalating operating costs have forced banks to diversify. Using IoT and machine learning to develop deep insights across vast data sets – and to create new, additional fee-based and wealth products targeted to the specific needs of individual customers.
Others have developed partnership ecosystems to increase value to customers through cashback, free insurance, incentives or loyalty programs like Avios. In the back office, work is being done to reduce the cost of acquisition and service using automation and robotics to streamline operations.
Digital giants and challenger banks are not hindered by legacy technologies and technical debt which incumbents often pass to the customer. Lower overheads and modern infrastructure allow them to offer a superior customer experience, personalised interest rates and premiums. They are also better able to accurately predict client propensity and demand, resulting in better capital and resource allocation to maximise customer lifetime value.
Multichannel operations are born
As high street branches are modernised for the demands of the 21st Century, contact centres have taken up the slack. However, this means that the contact team with training possess the skills and talent of the future for the front office.
Moving forward, financial services providers will need new data driven and analytical multi-skill disciplines to help customers. Further investment and transformations in contact centres will be critical to bridging the branch network through digital channels and SMART routing.
At the same time, there is a wide range of AI and ML underpinning operations from predictive demand forecasting, Natural Language Processing (NLP) and script-to-translate for Chatbots and contact centre agents. This is complemented by computer vision techniques for fraud and sentiment analysis.
Technology has expedited change
The use of technology has resulted in accessibility dependent on local digital infrastructure. Meanwhile, space-time compression has changed channel demand and the products and services available. It has also helped to deliver a better customer service in line with customer demands for more digital services.
In the past, corporate understanding of the customer and their needs and preferences was limited to their interactions with the staff in the local branch. With ML techniques along with technology tracking behaviour online and in digitised branches, it is now possible to develop a 360º picture of each individual. Despite these advances, most still lack an accurate way to predict customer propensity or to deliver personalised offers that align to the customer lifecycle value.
When deployed correctly, machine learning is helping to close these gaps. Personalisation and intelligent offers are easier to formulate and target, delivering optimised offers based on channel, device and next-best discounts.
Bringing digital back to the high street
The demand for personalised products and services means that branches can no longer simply be an impersonal outpost of the bank. Instead they must become part of the community, combining physical presence with technology – phydigital.
This involves using AI and ML technology to create a human-centred AI experience that caters to local demographics. Product offerings and services, even the layout and branch type, should be aligned to the needs of the local socio-economic population.
NatWest has already begun opening digital branches, and HSBC is reshaping for this new phydigital reality. Similar efforts by Virgin Money have realised a 200% uplift in sales at branches local to nearby Virgin casual lounges as a space for the community to use.
The omnichannel approach is increasingly important because branches are still important – at least from the customer’s perspective. People still intrinsically use branches for the human touch, particularly for complex activities like mortgage applications and reviews.
Omnichannel technology in a phydigital branch allows banks to better balance customer demands with their own interests. Arranging an in-branch video conference with a mortgage advisor gives customers the personal touch they expect – and allows the advisor to maximise productivity because they don’t have to travel to the physical branch.
Moving from multichannel to omnichannel
Banks typically have much of the data they need to improve products and services – but it is held in various silos. Information is neither democratised nor optimised for the realities of omnichannel delivery.
Most banks are investing in data science teams capable of turning data into insights and to better inform decision making.
The focus of these efforts should be on the preservation of capital aligned to customer VaR, customer shifts to digital first channels and hyper-personalisation. Consider the questions a customer asks themselves – How do I want to bank? What Channel? How do I want to buy and consume products? What is the most attractive rate? The bank can then begin planning how to serve these products and services to customers before they know they need them.
Integrating machine learning operations
Machine Learning Operations (MLOps) enable you to manage, monitor and optimise your machine learning lifecycle to maximise business value with ever-changing requirements, as well as, capacity optimisation for scarce talent pools. You can increase agile operations and build cross-functional teams better suited to each project.
Effective MLOps requires increased flexibility to integrate with other systems and teams to yield maximum value from your data. More important still is the cultural shift; leading from the top to focus on insight-driven operations that deliver a better service for customers.
Many banks are already investing heavily in advanced technologies like AI and Edge to drive change and increase efficiency across customer journeys. However, a relentless focus on the technology itself means that they are under-investing in the operational teams required to ensure sustainable value is delivered. Worse still, MLOps is often overlooked until scale becomes a challenge.
Enabling the MLOps transition
The Seldon toolkit can support clients at any stage of their MLOps maturity, growing and scaling as their requirements change. Seldon’s flexible architecture complements your existing infrastructure, frameworks and toolkits. It is agnostic and can adapt with the changing technical and business architecture in months or years to come, future-proofing value.
Omnichannel is a mission critical strategy for most financial institutions. The pace of change requires a flexible and collaborative platform – like Seldon – to ensure agile change to drive NPS and new business productivity/revenue.
Using Seldon you can enable a structural shift to a platform operating model, giving clients greater collaboration and control – particularly DevOps, data scientists and the wider business. Applying the same standards also builds greater transparency, auditability and agility. You will be able to leverage existing ways of working and integrate with the platform governance for each entity and use case.
Most importantly, you will have the data and frameworks required to deliver a true personalised omnichannel experience for customers.
To learn more about Seldon, MLOps and how to reduce the cost of servicing clients in the financial services sector, please get in touch.