Since the 2008 meltdown, banks have been arming themselves with strategies to counter a drop in revenue across most of their prominent lines of businesses and products. Recent Mckinsey report on Global Banking outlines a fall in revenue across business lines such as consumer lending & payments as well as a fall in sources of funds such as checking accounts and deposits, ranging between 5 and 25%. The revenue drop has been further accentuated by the advent of Fintechs and their patronage by the most progressive central banks of the world.

It is clear that the move towards better returns need to go beyond the conventional means of cost reduction and margin improvement. While the new age technology transformations like digital and analytics have been at the forefront of this shift, of late we are also seeing substantial traction in adoption of Internet of Things (IoT). IoT which comprises of a stream of technologies like machine learning, artificial intelligence, robotics and sensor based platforms is seeing a prominent uptake, primarily because of its state of readiness for deployment. Few of the instances have been captured below.

Cost optimization: Through a number of AI related initiatives, banks across the world have been reducing costs on the operations front across customer contact centres covering online, web or mobile, voice- IVR, call centres and branches. Banks like Santander, Mizuho Financial Group and Swedbank have introduced humanoid robots and AI based customer service assistance initiatives to bring about a change in customer servicing and experience. On the payments side, similar exercise is on using natural language processing, algorithms for matching and sensing and robots for doing the correction job.

Regulatory compliance: Post 2008, coverage and amendments to regulations have acquired a new dimension. This has led to the rise of automation of most legislations in Regtech – a branch of Fintech, through natural language processing, algorithms for matching & sensing, and robots for doing the necessary corrections. Case in point would be a prominent retail bank in UK that was successful in identifying fraudulent transactions with more than 90 percent accuracy, using machine learning algorithms. To take another example, machine learning on transaction data was deployed by a large payment processor to identify “mule accounts” involved in money laundering.

Credit decisioning: This space is marked by two engagement modes - bigger banks partnering with Fintechs deploying better algorithms, and banks improving their credit decisioning infrastructure through a dedicated internal team. Each of these streams have their unique IoT based technology adoption path. On one hand we have banks like Santander that partner with Kabbage, a small business loan provider that uses machine learning to predict bad loans, automate credit decisions and build credit risk models; on the other we have few new age banks that get a finer view of customers, their consumption patterns, lifestyles and associated risks through mobile applications, and accordingly factor the same into loan repayment calculations and increase profitability. What we see in these cases is data from mobile applications, sensors and IoT being used as inputs in credit scoring as a proxy to measure willingness and ability to repay a loan.

Customer acquisition and retention: Sensor data from customer devices, call records and purchase records are used to get insight on customers across aspects like location and movement, financial and economic background, identity and demographics, social behavior, usage and sentiment, all of which serve as proxies to build financial profiles for people. Incorporating these insights into analysis and decision-making allows banks to better understand its customers at an individual level and target a specific category. Prominent banks in US have been able to use this information to create product recommendations, design focused marketing campaigns and personalized communication streams, thereby targeting a specific category of customers through well designed financial products. Through deployment of IoT technologies, many banks have also enhanced their ability to predict when a customer is about to make a big purchase by incorporating clickstream data with purchase histories and behavioral patterns, thereby increasing conversion rates by nearly 25%.

While the examples above serve as a broad indicator of the role played by IoT in banking landscape, what clearly comes out is that early adopters have seen better results with these technologies than at any time in the past. The leap in adoption of IoT is not a region based phenomena, rather it has been adopted by banks that are true experimenters. Such banks are also better geared to handle the pressure on margins more effectively. IOT is not arriving, it has, in fact, arrived and is ready to invade the bastion of manual processing.