The term fintech is all over the business media. One does not really know when exactly it came into such ubiquitous usage. Separating the real power of fintech and the hype around it is a million-dollar task.
As I understand, fintech simply means using technology to deliver or enable financial transactions better. And frankly, innovations in this space have been going on in the traditional banking and other financial sectors for many years, if not decades.
The rise of retail financial services industry meant that millions of accounts, loans, credit cards, transactions, and payments were happening through banks, non-banking finance companies (NBFCs), asset management companies (AMCs), and insurance companies. Given the enormous numbers involved, it became necessary that technology be used to deliver the products and services to the customers. This was the low-hanging fruit and companies duly executed this in the past two decades through automation: point of sale machines, ATMs, IVRs, cards, mobile and internet banking, and others. This brought in operational efficiencies and, therefore, lower costs.
However, the ‘gold’ to be found in this space was not operational but strategic. Not just automation but intelligence. Technology can give incredible insights into how customers behave, what they buy, when they buy, how they repay, how they save and spend and a lot more. To extract this ‘gold’, fintech ventures came up with a multidisciplinary approach marrying marketing, technology, data analytics and product or value proposition design.
The crucial element is the data and information gathered about the customer and her financial life. This can come through:
a) Data at the point of sale, say, from application forms (demographic, psychographic), or previous transaction history, and
b) Usage and behaviour patterns after the customer uses the product (for example, current or saving account, loan or insurance plan).
Fintech companies try to use this data to make better offers, better targeting, better credit underwriting and so on. Fintech can exist as a solo venture or can be incubated within an existing, large business.
Across the world, fintech trends are at a nascent stage. The biggest hurdle for fintech to become intelligent is the integrity of the data going into the models that are built to form hypothesis, rules and strategies.
GIGO is a computer term for ‘garbage in, garbage out’. The GIGO factor is at play at the moment and, therefore, in my opinion, the fintech’s promise on paper may not get delivered in practice. Of course, this is a journey for many businesses and the space will evolve. Let us examine why the GIGO factor is a big hurdle to cross.
The first step in building models and algorithms is to rely on past data about existing customers or data in the industry, if available. In India, past data in most cases is not available or is incomplete or of poor quality. There are many factors that pollute the data and the information collected. Here are some of these.
Proliferation of agents: In India, often customers do not deal directly with the fund house, bank or the lending company or the insurer. Agents fill the forms, do the paperwork for the customer and therefore influence the data that is getting collected. In my experience, for example, agents influence and direct what data should be filled in ‘self declaration’ boxes, or which contact information should be given—it is not uncommon to find 30-40% false telephone numbers or emails. Even when companies provide direct web portals and mobile apps for their customers, there are instances where people give their user ID and passwords to their agents to operate.
Lack of professionalism and discipline of sales and operations staff: There is a culture of shoddy work at branches, or sales offices of most companies. The staff skills and habits are poor with respect to collecting data and recording it accurately on the customer relationship management (CRM) system. This is in contrast with other countries in rest of Asia, where the skills and discipline levels are higher.
Not enough modelling and testing in a complex market: Fintech will come into play only when robust and proven models impact the company’s revenue, product and service strategy. For example, one may conclude from behavioural patterns of consumers that they prefer a particular product feature, say, a payment term or a mutual fund type. But within that, there may be layers of variance hidden. For instance, people in Kerala may show completely opposite behaviour to that of people in Maharashtra. Averages in a market like India can be misleading. Full distribution ranges, or clusters, need to be peeled open and robust tests need to be done to establish strategies.
Lack of processes, tools and systems: Traditionally, the financial services industry is a laggard in investing in adequate systems and tools. Hence, the gathering of data is cumbersome, absent or partial. This hampers the fintech venture to harness the full power of its proposition. Even fintechs that rely on mobile or Web platforms are underinvested in terms of analytics. One reason for this is that many times fintech ventures are founded or designed by techies rather than by people with serious real-world experience.
In conclusion, the GIGO factor is a crucial factor to overcome if many of the current fintech ventures are to become successful.
Rajiv Jamkhedkar is founder and managing director of Serengeti Ventures Pvt. Ltd.