The initial phase of digital lending was characterized by entry of new fintech companies promising a disruption of traditional lending markets through fundamental innovation in customer experience, processing speed and risk management. Since then, banks and traditional lenders globally have recognized the need to evolve. As they migrate, what lessons can be distilled about digital lending?
"A successful digital lending organization needs to be a learning organization. It continually has to experiment and have the capability to learn from past mistakes"
We have identified five bedrock principles essential for banks and other financial services companies to drive excellence in customer experience and create sustainable profitable growth. Importantly, these principles are relevant for all types of credit – consumer, small business, commercial, real estate lending and a variety of credit products such as term loans, lines of credit, merchant financing, etc. The key principles for digital lending are:
1. Recognize that data and decision analytics are core to success
It is impressive to see the rapidity with which the expression “data is the new oil” has gained currency. There is good reason. An unprecedented amount of digital data is being gathered from an ever-expanding set of sources, including smart phones, browsers, apps, electronic files, digital cameras, accounting software, call centers, social media and satellites. Data from wearables, drones, autonomous vehicles, home voice assistants and other internet-enabled devices are not far behind.
Indeed, a senior bank executive of a midsized bank mused recently, “We need to be better prepared to deal with the next 10 years of data.”
Digital lending can take advantage of the torrent of new types of data that have become available and the new analytic techniques from machine learning and AI required to avail of them.
First, and especially important, is to use the data responsibly ensuring that there is customer consent, both implied and written. Then, there is need for a platform that can effectively ingest data from multiple sources and flexible enough to incorporate new sources. Lastly, functional expertise is critical to ensure that technology and human intuition are effectively leveraged.
There are several challenges confronted by traditional lending organizations in these respects which need to be overcome. Too much data itself can be a burden (e.g. a data landfill)… the value is derived from intelligent use which requires investment in talent and technology.
2. Maximize customer experience and risk control simultaneously rather than making trade-offs
Traditional lenders have historically navigated a tricky divide: they can offer either customer convenience, especially in terms of instant decisioning or exercise prudent credit and fraud risk management but have not been able to do both simultaneously.
This trade-off was made painfully apparent during the subprime housing crisis of 2008 and its aftermath. By simply putting their signature on a piece of paper, customers could arrange for a home loan without verification of income, employment and assets. This unrivaled level of convenience afforded to customers almost brought down the entire global financial system!
Since then, banks have undertaken a far more disciplined and careful examination of borrower creditworthiness. Except in the simplest cases, such an examination is often distressingly manual, time-consuming and expensive. It is also increasingly at odds with consumer expectation in the new digital economy, where they can shop for products, summon a cab, order food or arrange accommodation in a new city entirely through mobile phones and tablets.
By adopting mobile and digital approaches, banks have begun to resolve this fundamental trade-off. Customers can be on-boarded through an online loan application, either remotely or in a branch. The questions asked are relevant to the loan and the whole experience is paperless. The customer can give their electronic consent to KYC, income and asset verifications, and these can be conducted automatically through third-party APIs.
Other relevant information possessed by the customer can be digitized, analyzed, cross-referenced and assessed for rigorous underwriting, as well as for downstream collection management and cross sell, all in real time.
3. Use digital processes to enable end-to-end risk management
Traditionally, risk management has been a standalone function arising from handoffs in other parts of the customer cycle. At the same time, every activity within the financial institution impacts the risk, ranging from origination, technology, underwriting, customer management, cross-selling, servicing and collection.
Digital ensures that each function impacting portfolio credit risk leverages all available data and technology for excellence in risk management across the entire life cycle. The key is that a digital lending organization captures information in a scalable way and makes it broadly accessible for both customers who have received loans and also their repayment track records.
This can lead to better customer scorecards which can be distributed across the organization through digital workflows. Marketing can monitor credit quality by channel and help make changes in real time. Similarly, sales can prioritize customers by scorecard metrics for outreach and cross sell. Servicing, collections and even legal can similarly be impacted.
4. Build internal and external ecosystems to jump-start origination
A major priority for digital lending within traditional institutions is generating volume from the outset. While the growth rate of digital lending might be high, the fact that it initially has a small amount of volume means that it can be disadvantaged on managerial attention and resource allocation, limiting its future potential. For this reason, it is essential that managers plan how they intend to grow loan origination in parallel with the implementation of digital technology.
The first and most natural place to look is the existing customers of the financial services company. For example, within a retail bank, it could be deposit customers, wealth management customers, business banking customers or others who need lending products. E-mail campaigns, preapproval offers and technologies (such as API interconnects) that enable cross-selling can be prioritized.
The above is what can be referred to as an internal ecosystem. Digital capabilities also lend themselves to building powerful external ecosystems. These consist of partnerships with other companies, enabled by technology and data science. For example, a bank could launch new lending products with a payments processor, insurance company or other strategically important customer aggregator. Robust internal and external ecosystems can enable rapid volume growth following the launch of digital lending.
5. Transition the organization to a digital mindset
A common perception among managers in traditional financial services businesses is that digital is just one more channel among many. In fact, success within digital lending requires a completely different way of thinking.
Many traditional lending institutions simply put their existing business process for lending online, which often imposes a heavy informational burden on the customer. This typically results in a large drop off in customer applications. Instead, they need to design their customer journey to effectively manage the trade-off between customer convenience and risk management mentioned previously.
Digital lenders also need to learn how to execute continuously. Within small business lending, for example, more than 50% of applications arrive outside regular business hours or on weekends, requiring customer attention and underwriting on a 24/7 schedule.
A successful digital lending organization needs to be a learning organization. It continually has to experiment and have the capability to learn from past mistakes. Experimentation can be through A/B testing, both for in take of customer applications and also at the back-end with risk management. Each activity will generate data which can be further analyzed using AI tools.
Nowhere is the necessity for a new way of thinking more relevant than within risk management. One instance lies in what can be called credit forensics. This process consists of systematically analyzing historical underwriting decisions with the information available at the time and identifying factors that are predictive of subsequent defaults. Due to the ready availability of retrospective digital data, the level of actionable insight that can be achieved is striking and can substantially lower default risk.
More broadly, the risk function can be tasked with analyzing vast amounts of digital information and disseminating it efficiently and effectively across the organization. Risk managers, rather than being specialists who interact primarily with each other, now need to communicate extensively with non-specialists in other departments to ensure credit risk is properly managed throughout the organization to enable the goal of end-to-end risk management.