FinTech Banking on IoT

Smart Business

FinTech Banking on IoT

Smartphones, automation, and artificial intelligence are helping revolutionize how banks deliver services, but these technologies may also prove the undoing of the traditional players. Will today’s big banks retain market dominance, or will upstarts and apps fragment the banking sector?

by Stian Overdahl

Banking, one of the largest and most profitable industries in the world, is at a turning point. The 1,000 biggest lenders have combined as-sets worth around $123 trillion, with an average return on assets (ROA) of 0.9 percent. Despite all this heft, the future is uncertain as technology upends their industry. At the heart of this shift is the smart-phone. From the banks’ viewpoint, these ubiquitous IoT devices have provided direct benefits. Apps bring banks closer than ever to their customers, who can make payments or international money transfers via their phones. Previously, these transactions would have required them to visit a branch office, make a telephone call, or write a check. New technologies can also offer immense savings for banks, allowing them to close local branches and automate many worker functions.
With the rise in functionality and popularity of banking apps, desktop and laptop banking is stagnating and traditional telephone banking is dying out, while visits to branches are becoming less common, says Max Flötotto, a partner at McKinsey & Company. In the German market, for example, 20 percent of customers ac-count for 90 percent of branch visits. Increased smartphone usage generates new pools of customer data and this creates more sales opportunities for banks. “There are a lot more opportunities to create sales leads and to interact with customers if they look at their banking app four times a day than if they come to a branch every 12 weeks or once a year,” says Flötotto.

There are a lot more opportunities to interact with customers.
Max Flötotto, McKinsey & Company


Smartphones have also flung open the door to all kinds of competitors – challenger banks, big tech, and start-ups – all chasing after their piece of the financial world’s immense prof-its. In Europe, a wave of online-only banks, neobanks, such as Revolut, Monzo, and N26, have seen massive growth in customer accounts in just a few years. Part of their value proposition is lower fees – Revolut, for example, offers currency transfers at the inter-bank rate – but much of their success can also be attributed to a superior user experience, through slick apps, utilitarian sign-up processes, and showy marketing campaigns. Neobanks have also showed that banking relationships are far less “sticky” than they used to be. The old joke that a banking relationship of-ten lasts longer than a marriage no longer rings true. To date, the biggest growth by fin-techs has been in emerging markets where many consumers are un-banked, or under-banked, and hungry for new digital services. In the Asian markets, growth of digital payments has risen sharply, epitomized by Alibaba spin-off Alipay and Tencent’s WeChat in China, or the fast growth of mobility apps, such as Singapore’s Grab, that have built-in digital wallets. These apps have further squeezed banks by diversifying into related sectors, including products for the credit market, investment services, and insurance. Will a similar upheaval take place in developed markets like Europe and the US? For all the marketing savvy of neobanks in developed markets, many are only used as a second account by customers still choosing to pay their salaries into traditional bank accounts. Some neobanks still lack core banking products such as personal finance or mortgages (and the familiar questions about the profitability of well-funded start-ups lingers in the background).

Personalization of Services

A new Look at Existing Clients

Chatbots are increasingly being used as a customer service tool by banks, especially online-only neobanks. As in any industry, AI-powered chatbots can provide in-formation to customers, allowing banks to employ fewer call center workers. AI can also be used to comb through a bank’s customer base to look for miscategorized, high-value clients, or to identify appropriate clients for new products, such as savings plans.

Many experts believe that change in developed markets comes at a slower clip. In China, it was the failure of retail banks to effectively meet the needs of a growing and increasingly wealthy middle class that set the stage for the rapid growth of digital payments, says Kevin Kilty, the CEO of Hubpay, a digital wallet service targeting Asian workers in the Gulf states. He sees a possibility for rapid growth of fintech in other emerging markets but is doubtful that financial services in developed markets will be similarly disrupted, given that most Western banks already offer consumers efficient payment technologies – such as tap and pay on their credit and debit cards.

Failure to meet the needs of a growing middle-class set the stage for the growth of digital payments.
Kevin Kilty, CEO HubPay


That said, the steady growth of fin-techs and shifting consumer behavior may contribute to a steady erosion of profits for banks in developed markets. “It’s a critical time [for banks] but I don’t think it will be an overnight shift to a new model,” says Flötotto.
Philipp Baecker, expert vice president at Bain & Company, says that banks in developed markets do have a battle on their hands – and a crucial part of that is about ownership of the customer relationship. The worst-case scenario for retail banks is that fintech and tech companies become so successful in offering financial services through their own platforms that banks are relegated to a back-end role, processing products for tech players that control the customer relationships, he says.
“What makes this so fundamental is that in the future a lot of the services offered today that we know as financial services will be integral parts of a customer experience that may cut across industries. The question is: what is the role of banks and how visible will they be to the consumer?” Owning the customer relationship is not only important from a brand and revenue perspective – it also means potentially owning customer data, which itself is increasingly valuable.
FinTech Banking on IoT

Hipness Counts

While banks can have immense profits, they also have colossal costs JPMorgan Chase, the largest bank in the United States, has been spending around $11 billion per year on technology, while Spain’s Santander Group announced, in 2019, that it would spend €20 billion over four years on its digital transformation. Reducing costs via technology, specifically automation, is therefore a major opportunity for legacy banks.

What is the role of banks and how visible will they be to the consumer?
Philipp Baecker, Bain & Company


However, it’s not as simple as spending money to get results. Experts caution that one of the biggest obstacles that banks face is legacy mindsets among managers and staff, with a tendency for bank projects to over-run budgets and schedules. Moving bank staff into a start-up-style hipster office won’t have any effect if they bring with them traditional working styles, such as a strong aversion to failure that limits their willingness to take risks. Still, the restrictions that arose from Covid-19 and the need for contact-less banking have “accelerated need for banks to digitize massively and, in some cases, made what was believed impossible to be possible within the shortest time frame,” says Dave Murphy, a managing partner at business transformation specialist Publicis Sapient. Companies have become willing to implement digitization measures they have been delaying for years within just weeks. Whether this will result in a permanent shift remains to be seen.

Credit Scoring

Keeping Score by Smartphone

Evaluating credit risk is one of the most fundamental activities for a bank, and AI is increasingly being used to gauge credit risk. In developed countries, traditional credit scoring companies with large data sets are using AI to score customers more accurately. In emerging countries, where some first-time borrowers have no credit history, different methods are used, such as using smartphone metadata or social media information to create a credit score. Singapore’s monetary authority has even launched a framework to ensure fairness in AI credit scoring methods.

FinTech Banking on IoT
Neobanks, meanwhile, are well-positioned to take advantage of new technologies in banking, including big data, cloud services, robotic process automation (RPA), and artificial intelligence (AI). Newly minted banks are often able to integrate these in their systems from day one, rather than struggling to transform or merge outdated legacy systems, a thorn in the side of many banks which not only affects the adoption of new, data-driven tools but also inhibits the ability to control costs via digitization. Sarah Kocianski, head of research at 11:FS, a UK-based financial consultancy, says banks can face a raft of challenges, such as customer data siloed within individual systems (between a checking and credit card account, for example) – and even data that hasn’t yet been digitized.

Covid-19 has accelerated the need for banks to digitize.
Dave Murphy, Publicis Sapient


By contrast, tech companies and neobanks are more likely to have managed their customer data from the start to ensure it’s clean, efficiently stored, and easy to access in data lakes. “Many banks didn’t react early or quickly enough to the ideas of how important data could be, to ensure that they were a good place to use the data they had,” she says. Not only can this hamper their implementation, it can also make banks less attractive as an employer for AI specialists because they don’t have clean data sets to work with, Kocianski adds. This is significant as the development of tech capabilities in the financial world is as much a battle for talent as for anything else.
While RPA can be used to automate existing processes in many banks, total redesign of a bank’s systems may be required to gain full advantage of new technologies, especially AI, says Bain’s Baecker. “You can’t simply take your existing processes, your existing products, existing service models, and your existing mindset, and put AI on top,” he says. “That is the edge many of these new challenger banks and even larger technology companies have – they don’t have to deal with as much legacy and can rethink everything from scratch.”
FinTech Banking on IoT

From RPA to AI

Michael Berns, director of AI and fin-tech at PwC, says that many of the European neobanks have been built on microservices with RPA included, while some of them have AI components as well. Revolut, founded in 2015, is UK-based and now has almost four million active users each month. It has built a chatbot for viral marketing which it is using to good effect to generate additional customer contacts, says Berns. “I doubt they would have been able to do that in a scalable way without fully leveraging automation, RPA, and AI,” he says.
Berns sees the beginnings of AI in banking as a response to the series of massive fines slapped on banks following the financial crisis for a host of wrongdoings, including money laundering and market manipulation. Prominently, HSBC received a $1.92 billion fine in 2012 for failing to act on suspicious transactions going through its books, many of which were related to Mexican drug cartels. Scandals such as this put US banking licenses for a number of European banks under threat.

Many banks didn’t react quickly enough to how important data is.
Sarah Kocianski, Head of Research, 11:FS


“Banks had little choice but to resort to AI, to fulfill the urgent requirement for new methods and complex tools to prevent further massive regulatory fines,” says Berns. “AI is a key tool in these areas to analyze communication and gain more control over internal proceedings.
“The aftermath of the financial crisis has seen the use of AI spread and it is now used in areas including the detection of payment fraud, suspicious transaction reporting, credit scoring, contract analytics, and a host of applications around lead generation and customer-service personalization and response. A classic use case is the combing through customer data to look for high-value clients that a bank might unknowingly have on its books.
FinTech Banking on IoT

Banks had little choice but to resort to ai to prevent massive regulatory fines.
Michael Berns, Head of Research, 11:FS


Behavior such as dining out at expensive restaurants in London and Paris might be a tip hat the bank has an affluent client, perhaps using a secondary account. Identifying such a client within a bank’s existing customer set is far more cost-effective than acquiring new customers in this segment on the open market, says Flötotto at McKinsey’s. One company providing white-label AI solutions to banks is Personetics Technologies. Dorel Blitz, the Israeli company’s vice president of strategy and business development, de-scribes its product as an “AI-based proactive and personalized engagement platform that is designed to help banks to better analyze their customers’ financial behavior in real time at a very granular level.” Personetics can be used by banks to engage with their customers through a digital or physical sales channel, for example to suggest that customers take advantage of an automated savings plan. “The problem that we are trying to solve is helping banks to stay relevant, to stay in the center of their customers’ financial lives,” he explains. Personetics’ goal is to help banks fully personalize their offerings to customers, in real time. Blitz believes banks will shift from the current passive approach to a fully proactive one and evolve into having “full automated money management capabilities, where your bank will be able to think and act for you, instead of just nudging you and telling you where you should actually put some money aside”.
FinTech Banking on IoT

We’re able to help banks to increase their data enrichment and better understand their customers.
Dorel Blitz, Personetics Technologies


The platform can work with any kind of structured data available to the bank and, while it’s typically focused on core banking data, such as checking deposits, loans, and credit cards, it can also incorporate external data the bank is collecting, such as geolocation, or rewards and loyal-ties data. “We’re able to help banks to dramatically increase the data enrichment and categorization to have a much better understanding of their customers’ financial behavior,” Blitz says. Currently Personetics’ solution doesn’t make use of social media data (though as a white-label solution it can make use of any structured data, notes Blitz), but this is a growing area of focus in some markets – and also a controversial one. Differing attitudes around the world as to how companies use customer data, and political realities, are expected to result in significant differences in how AI is adopted, given the need for training, validation, and test data sets to develop algorithms. Take for example AI-powered credit scoring. Though AI is used to enhance traditional scoring methods, in an under-banked market where the applicant doesn’t have a financial history, an alternative solution is to scrape personal data from social media to build a credit score. Even though some applicants willingly give permission for an algorithm to scan all their activity on social media, the concept doesn’t sit well with everyone.
FinTech Banking on IoT

Anti-Money Laundering

Learning to Spot Suspicious Signs

With some of the world’s biggest banks having received multi-billion-dollar fines for money laundering, anti-money laundering (AML) compliance is a significant cost for the industry. Machine learning techniques can help banks sort through transaction data for signs of suspicious activity. Nevertheless, adoption of AI technologies for AML in North America has been relatively slow and is often limited to proof of concept, according to a report by Accenture. One downside is the notion of machine learning as a “black box” where the inner workings are not understood by compliance officers or the regulators who require them to understand and validate how AML outcomes are derived.

FinTech Banking on IoT
Attitudes to privacy tend to be cultural and, as a result, what is permitted varies greatly between regions. In much of Asia it is permitted to scrape social media and then come up with a credit score or sell products, such as insurance, based on this. But in Germany, for instance, it seems less likely for the regulator to accept AI factors as part of a credit-scoring model. “The focus here is on transparency as well as explainability,” says Berns. In particular, with China reckoned to be a global leader in AI, the fact that citizens’ personal data is routinely collected by the state is seen by many experts as likely to give it a further edge, as Chinese consumers are more likely to be willing to give up personal data if they see an ad-vantage in doing so. “In the end, it is a global race and the winners are the ones who have the most diverse data set. From a maturity perspective in terms of removal of blockers, it certainly helps if the government and regulations are fully aligned with regard to col-lection and use of data,” concludes Berns.
FinTech Banking on IoT

A Means Not an End

The overall business environment for banks is also likely to play a role in determining how aggressively banks are willing to make big investments in tech like AI. On this front, profitability of US banks is much higher than European banks. When it comes to investment decisions, size matters, too. Ron Shevlin, director of research at Cornerstone Advisors in Boston, Massachusetts, says that, while he is bullish on AI, in the longer term the reality is that thousands of smaller banks will be unable to develop their own technology and will rely on vendors.
FinTech Banking on IoT

Banks need strong strategic approaches around busi-ness, data, and vendors.
Dorel Blitz, Personetics Technologies


Picking a vendor that fails to keep up with developments could be costly, he says. Shevlin says he is fond of pricking the “AI hype bubble”: “Too often AI is talked about as if it’s a technology that’s going to solve all problems. What it really comes down to is that [banks] need strong strategic approaches around business strategy, data strategy, and vendor strategy.
“That view is echoed by Blitz at Personetics, who cautions that for a bank to effectively use AI it must first have clear notion of its business goals and strategy: “AI is just a means, it’s not the goal.”
FinTech Banking on IoT

Leave a Reply

Your email address will not be published. Required fields are marked *