IBS Journal: The iconic monthly FinTech magazine

February 2020 issue out now! Subscribe now

Launching India FinTech Report 2020

Market sizing and forecast of Digital Lending, Payments, WealthTech & RegTech. Subscribe now

Leading Back Office Systems for Banks

In-depth Supplier Profiles and User Lists. Subscribe now

In-depth Banking Tech and FinTech Research

Over 300 research reports that's updated quarterly. Subscribe now

Analysing the future

Retail banks often find it hard to master and analyse customer data across various different brands and silos, from structured and unstructured sources, especially in comparison to neo banks, retailers and others, but Neil Ainger finds a determination to improve and hope in AI and cloud computing for the future   

The ability to collect, slice and dice data has been around for many years. It is common for bricks and mortar, online and mobile retailers in developed markets in the US, Europe and Asia, for instance, to squeeze every bit of relevancy from every piece of information that the customer gives to them, including on social media and regardless of the channel, so that log-ons and onboarding is made easier and to ensure that cross or upselling opportunities are not missed. Internal efficiencies and easier compliance can also accrue from better use of data and the elimination of silos, while brand loyalty and value are improved.

“There are lessons a retail bank can learn from other industries and from some of the newer companies in financial services,” says A. Charles Thomas, CDO at Wells Fargo, pictured, while stressing the advantage these newer FS companies enjoy having “started out with a technological blank slate”.

“I don’t think traditional banks are ‘lagging behind’ in their mastery or use of data.” he adds, emphasising the need for relevancy, bridging departmental barriers and internal cultural change. “It’s just that the best of us are playing a long game and have a keen focus on working within our existing data ecosystems, while keeping a critical eye on when to leapfrog to new investments.”

Those new investments might include artificial intelligence (AI) powered chatbots, robo-advisors and machine learning tools; Hadoop for unstructured Big Data analysis; and cloud computing, which increases bandwidth and technical data sharing capabilities and is especially useful when allied to the coming world of ‘open banking’ and application programming interfaces (APIs). The UK Competition and Markets Authority (CMA) want to see Open Banking by 2018 and it’s inherent in the EU Payment Services Directive (PSD) 2 regulation.

As Thomas’ colleague, Bipin Sahni, Head of Innovation R&D at Wells Fargo, explains “increasing the number of data sources and creating a holistic view of our customers is key” to the future success of the bank. Our focus is on how data can improve the user experience. Enhancements in AI are opening doors to endless opportunities. These engines can replace previously labour-intensive programming workloads, and allow us to incorporate many more data sources. For example, we’re using AI from both natural language processing (NLP) and analytics bases to build a framework, so that our business lines can learn from one another. In addition, Splice Machine, a company in our FinTech Startup Accelerator programme, is helping us discover data-driven insights to help drive predictions.”

Thomas cautions, however, that “the best technology in the world won’t matter if we don’t know what to do with it” and stresses the need to find the “right talent” and for the bank to “organise itself around enabling these highly skilled team members to do their thing” and “re-imagine how we work.”

FinTech ‘co-opetition’   

The threat from newcomers to the retail banking market, such as Atom, Fidor, OakNorth and so on is real, but it is often targeted at niche customer segments such as Millennials or entrepreneurs and it is small-scale at present, and reliant on traditional banks not responding aggressively. In fact, what most banks are doing is co-operating where they can with FinTechs and competing in other areas. It is a model that is often referred to, somewhat clumsily, as ‘co-opetition’.

“If customers flock to these so-called ‘challenger’ or tech-based ‘neo’ banks, which typically use data better as they do not have legacy IT or silos caused by past acquisitions, additional regulatory obligations and costs might hinder them as they approach Tier 2 or 3 status,” says Zhiwei Jiang, Global Head of FS Insights & Data at Capgemini. He doesn’t believe they’ll have a major impact in future years as the excitement around FinTech innovation recedes and tech lessons are learnt by incumbent banks, barring perhaps one or two standout startups that may infiltrate the market on a truly large scale.

As Szymon Mitoraj, Head of Digital at ING Poland, says: “You shouldn’t ignore FinTech startup innovation and clever data apps. But I only really fear the bank that cooperates with them better, not them as standalone entities.”

“It is strategy and end use applications that matter, not technology in and of itself,” adds Mitoraj, who was recently at FinovateEurope 2017 in London showing off his bank’s new MyING advisory and finance management mobile tool. This uses analytics to offer savings tips, targeted loan offerings, drop-down data filtering menu options and a search capability. Speaking exclusively to IBS Journal after his demo at the show, Mitoraj explained MyING has “already been used by two million customers in the last six months in Poland”, which is scale “FinTech firms can only dream about…There was a huge amount of data to consolidate but we were helped by the lack of legacy IT in our bank locally and Poland generally.”

Interestingly ING is also trying to re-enter the UK retail bank market from its Dutch home base via the Yolt app. This mines customer account data across different bank accounts – whether HSBC, Barclays or wherever – and then provides tips on budgeting and savings. Yolt can also predict balances, based on standing orders and direct debits, and has various alerts for overdrafts, due payments, spending pattern analytics and so forth. It effectively acts as an aggregator – a money supermarket if you like – and could offer ING a FinTech route back into the UK market it left five years ago.

Yolt is an example of how open APIs and banking – encouraged by the EU’s PSD2 and other such regulatory initiatives, not to mention recent technological advances – may disrupt the marketplace. However, banks will naturally look to develop their own responses and to make their own service and products good enough to dissuade people from switching.

A flood of customer account data will result from open APIs, which could be used by secure newcomers, challenger banks, FinTechs and so on to shake up the marketplace, as happened in the insurance market where aggregators initially flourished. The long-term success, or otherwise, of retail bank aggregators offering alternative services will depend upon how the banks’ respond.

According to Capgemini’s Jiang it is large scale Asian, Latin American and Middle Eastern banks, with newer IT and “less legacy drag” that are likely to be the true competition and/or template for established Western banks to follow, not neo banks that lack scale. He thinks the best banks need to do three things to help improve their customer data analytics: Adopt a mobile channel first approach; Look at the customer journey, not the process; Introduce a new core banking system, or truly digitally transform it.

“The core isn’t sexy, exciting or as easy to get VC funding for, unlike many front-end FinTech developments,” says Jiang. “It is on the operational back-end of the business so it takes hard work to transform it. You need to innovate in the core, however, in order to improve customer data handling and analytics universally.”

The requirement is to be data-centric and aligned, moving away from a business versus IT approach, so that people, process and technology are aligned. “Most Tier 1 banks aren’t really changing,” says Jiang, barring a few exceptions. “They are just bolting on new apps, mobile channels and so on, without doing the necessary underlying transformation in their core banking systems, or changing culturally.”

Greater use of cloud computing or blockchain technology in future years could transform the core, or even wide-scale adoption of the Basel Committee’s BCBS 239 principles on risk data aggregation and reporting (RDAR). These 14 principles should flow into a Tier 1 retail bank’s operations and data-handling and analytic capabilities from its investment arm and risk department, but as they cover governance and technical data disciples they are notoriously hard to achieve.

The transition from one part of a large Tier 1 bank to another won’t happen automatically. The point is that banks must move towards being truly data-centric organisations, and away from a procedure-based approach, but this takes time.

“Traditional retail banks are very process orientated,” says Jiang. “However, new banks in Asia and elsewhere [with no legacy – Ed.] make no distinction between IT and the business. They are not so procedure-driven. If you take this data-centric approach you should be able to take information out of a cloud-based data lake.”

A cloud-based data lake might be the ideal, but as Jiang admits “it’s not easy” to achieve. According to David Wallace, Global FS Manager at SAS, banks face a number of obstacles in turning the often-siloed customer data they already possess – plus the external unstructured data that is available – into a world-class customer experience.

“Banks have to break through the ‘silo fog’ to get to a single, constantly evolving digital picture of each customer across every relationship,” he says.  “They must undergo a digital transformation, in which internal silos are taken apart, moved out and replaced by a single vertical stack of completely integrated applications that support mobile banking and are enabled by advanced data analytics. The latter will provide the insight, agility and flexibility needed to understand and serve customers in the same way they see in the rest of their digital lives.”

Todd Winship, Temenos’ Product Director for Data & Analytics, believes that: “If a bank wants to provide a true 360-degree view of the customer, they must ensure they can achieve accurate matching between core banking and external systems.” Blending internal structured and external unstructured data is important.

Cross-departmental capabilities are also crucial. A fraud department at a traditional retail bank might be excellent but its information is not being fully utilised unless it is shared securely. “Likewise, a compliance department is typically risk-based and siloed,” says Björn Holmthorsson, CTO at five degrees, a Dutch-based digital banking platform provider, who was formerly CIO at Landsbanki Luxembourg.

“At new challenger banks, such as Knab in the Netherlands where five degrees provided some of the technology, they are trying to make compliance part of their business process, and thereby more efficient. They are aligning people, process and technology better, so that onboarding, efficiency and customer service is easier. The key advantage of challenger banks is they are ‘greenfield’.”

Knab already has 125,000 customers in the Netherlands, particularly among small-to-medium-sized enterprises (SMEs) and retail bank customers switching from incumbents. It has only 120 staff, no branches and low overheads. “Such banks are definitely a threat,” says Holmthorsson, “and an encouragement for traditional banks to try a new way.”

The three main obstacles to achieving a data-centric bank in the opinion of Temenos’ Winship are:

Source data availability to core systems: “All of a banks source systems must be able to provide robust data, with business context, in real-time to achieve strong customer analytics,” says Temenos’ Winship. “Too often, however, critical data solutions such as core banking do not yet have robust data integration capabilities, such as the ability to omit event streams or provide extensive data replication. This results in the data being trapped in the source system and unavailable to analytical solutions.”

Data platform & unstructured capabilities: In terms of the data platform, Winship believes, “many banks have not yet invested in the next generation of Big Data platforms. Banks are currently running traditional data warehouse solutions that, whilst effective, cannot respond rapidly to real-time and unstructured data.”

Analytical skillsets & staff: Lastly, regarding skillsets, “banks struggle to hire and retain staff with deep analytical experience,” says Winship. “Banks must look at more innovative compensation models, partnering and other [FinTech – Ed.] collaboration techniques to ensure they have the skills required to utilise analytics most effectively.”

According to Boxley Llewellyn, VP of Watson FS Insight Solutions at IBM, which has experience of AI going back to its Deep Blue machine and is now focused on cognitive technologies, “the industry is just realising the importance of analytics to provide more valuable insights – concerning behavioural segmentation, cashflow predictions and life event predictions.”

“Banks can and will find ways to improve their customers’ experience via advanced analytics,” he says. “New entrants have begun to grab snippets of this value, but a bank has advantages too – such as well-organised pre-existing structured data; the ability to integrate third party data; increasingly cloud-delivered analytical options; and the scale to bring them to market faster.”



Related Posts