'We are a Very Data Driven Company': Bharath Thothadri, MD, RIMCE, American Express

Bharath Thothadri, MD, Risk & Information Management Center of Excellence, talks about American Express is leveraging customer data to server better and earn better. American Express 's Risk and Information Management CoE provides research, development and implementation of cutting edge risk and information management solutions for its worldwide businesses, working hand-in-hand with cross functional colleagues across multiple geographies. The mission of RIM is to drive profitable business growth by providing analytical insights and decision support to the Global Risk teams, the business LFO teams, Global Credit Administration and key partners.

What is American Express' philosophy, as an organization, about information and risk management ?
BT : We are a very data driven company. With over 10 million merchants and over 100 million card members we are sitting on data worth trillions of dollars across the globe.
Through their cards we have an incredible amount of information on who they are, where they live, and through their transactions we know what their preferences are. So our whole premise is how do we use technology to use this data for various aspects of our business. So right from prospecting to targeting with the right products and services to approval for credit to collections, we have to use this data. We have also started a number of businesses, and they are in early stages, where we try and leverage and monetize the data we have.

Any examples of these new businesses ?
BT: I can cite two examples. We have over 10 million merchants, and these merchants know quite a bit about their customers and what they do in their shops and stores. But these merchants have no idea what their customer do once outside their establishment, or what they do with their competitors. We have started this business
called Business Insights, which leverages all the data that we have, and its primarily based and driven out of India. The second example I would like to mention is about monetizing our data in a very different way. Called Performance Marketing, in this business we scale out the data we have and use advanced statistical models like /'look alike modelling'/ to identify potentially new customers at the global level who behave like my existing customer, and hence could become my customers. This is a very useful information for instance for ad tech companies as it helps predicts how are a broader set of individuals likely to behave.

How much have these technologies have helped AmEx or its customers in terms of enhancing revenues or profitability, or new revenue streams ?
BT : In our core area, which is the prediction of likely hood to payback, this is very useful. For instance, the two most important we need to check out when somebody is using our card is whether he or she is our genuine user and not somebody who is fraudulently using our card, and secondly, will the user pay us back. If you look at the credit matrix alone, then every year, in 1 basis poin timprovement in our credit modelling risk, will create an approximately USD 25 million profit impact for us. And we regularly improve our models by 10 to 100 basis points, so the annual impact just
because of credit risk models is in the tune of a billion dollars. And then there is fraud, which is much smaller, Plus there are these new revenue generation models. For a revenue base of about USD 30 billion, this is significant, and it is growing. So data is really the key about who we are.

How is the situation on the expertise availability ? Sometime back getting data scientists who know how to cull out the relevant data and read it were very difficult to get in India.**
**BT :* It is not easy at all and we work very had on it. As we are growing fast, and we also send out experts to other geographies, we hire about 150 people every year. Out of this about 70 come from campus, and another 30 from refer ralsfrom those recruited from the campuses; the remaining are recruited from the market. Our team works very closely with our HR team for campus recruitment to generate interest amongst students, as well as identify ones with the right interest, skills and aptitude. We were the first and perhaps the only one to launch an app for the campuses which students can download, and through the app we interact with the students on a continuous basis. There are special games for the IITs and IIMs through which students and
teams compete with each other, and build ideas and proposals. These activities give
the students a sense of what the job entails. This is helping us a lot because earlier people thought that even if it is American Express, it is most probably a back end kind of a job. But these activities have helped this job being looked as interesting, important and high profile. This year we have been a /'day zero company' /in all the IITs and IIMs, like Facebook and Google.

Do you sense that the demand for these data scientists is going to grow ?
BT : With so many MNCs, Indian companies, and start-ups looking for data scientists, every year it is getting harder. Data science is becoming a great place to be in and the demand is huge and growing. I am entirely convinced that artificial intelligence and deep learning for leveraging data to automate fairly advanced things, is now happening in a big way now. Just consider the billions of smartphones that come loaded with Android and Google which has personal assistance tools that, and imagine the amount of data that is being created.

Just like ICT has played transformational role in some sectors like taxi services, hotel booking, train tickets, banking etc, do you see something similar on the cards as far as credit cards or other financial services are concerned ?
*BT:  Absolutely, as you would know that fintech is the hottest thing in town, globally as well as in India. If you look at companies like Capital Floats or India Lends, that have build technology platform and are possible game changers in financial inclusion. They may or may not lend money, but they gather information on individuals and then use advanced algorithm models to asses their credit risk, and then different lenders can then bid for these customers. The important thing is that these companies are reaching out not to people like you and me, but a completely new population set, that had so far little access to financial services. And this will totally transform how borrowing and lending happens. These startups are building these tech platforms and might or might not lend money, but are aggressively gathering and analysing data. And most importantly they make it a very simple and seamless experience for the customers, in contrast to to the long and lengthy procedures in traditionalbanks that takes and weeks. Its just by sharing some basic information, a click of the button, and they can approve loan applications in a few seconds.

How do you see this space evolving in India ?
BT : I see quite a few forces at play here. There is quite a few private equity and venture capital money coming in. There are quite a few examples of similar successful stories
in other parts of the world, that can be emulated and customized for India. There is quite a bit of entrepreneurial action these days in India. The skill sets required for such
work is basically data science and software, and we have enough of that. And the government is also trying very hard to reduce friction in financial services, be it payments,  or lending and borrowing, and all this will be riding on Aadhaar. Its
still early days, but the vision is wonderful. India is a market where the demand for credit far exceeds supply and most of the people out of the 1.2 billion population do not have access to credit because the right processes and systems do not exist,
although they are able and willing to repay. My view is that this is is an amazing opportunity.

Gathering and using customer information is a very sensitive issue, yet it is being misused so blatantly these days. When there is so much data thrust for everything American Express is doing, does it create conflicts and dilemma for you all at the planning stage ?
BT : For us there is no conflict, because unlike many other companies, in the end the value of our brand is incredibly large, and the risk of dilution of our brand value is too big for us. It is a dilution of the trust people have in American Express and that is paramount for us. That is why we have a very comprehensive set of compliance and legal systems and processes with regards to gathering and using information. Our disclosures are very clear and transparent.

Today, American Express is seen as very exclusive. Its cards are perceived as very elite, and there are many establishments that do not accept Amex cards. All this technology deployment that Amex has embarked on, will see Amex cards and services for the common man ?
BT : Absolutely. We are dramatically increasing, globally as well as in India our merchant coverage. On the acceptance side, there is a global initiative OptBlue where we are leveraging third party companies to sell our acceptance to merchants. In India, we have partnered with ICICI and SBI for their merchant base. On the issuance side, in India we have both proprietary cards as well those issued through our network
partners, and both of them are growing rapidly here . In fact, for our proprietary cards, India is one of the fastest growing markets globally. Wherever appropriate, we are looking at newer sets and groups, to get American Express cards are in a broader set of hands. For instance, we are developing credit bureau capabilities for those who never had credit card before so that their credit worthiness can be quickly assessed through their bank account information. We are constantly and aggressively using data to profitably underwrite more and more people. With the help of data we are also introducing, where appropriate, a more diverse set of products. Thus we have been able to come out with card for consumer segments whose needs are totally different. For instance for those who travel a lot we have a card along with Jet Airways, and
another one with Make my trip. With all this, we have been able to significantly increase the average spend of an Amex card user. As per RBI data, an Amex card member spends five times more than other card users in the market.