What Eliminating noise did to this Data Analytics project at HDFC Securities

Last Updated: Fri, Oct 25, 2019 17:44 hrs
HDFC Securities

Data in a digitized world is "the oil". But does that mean organizations cram petabytes of consumer data on their analytical engines to develop better cross-selling opportunities? Running any analytical project from such a huge repository of data may sound alluring, however it may not have the desired benefits.

In the words of NK Purohit, Head Digital Strategy and Analytics at HDFC Securities, organizations can reap great benefits by simply deciding on what to drive from their analytical projects rather than using analytics to crystal-gaze and getting bombarded with a clutch of questions.



Purohit's assertions are credible considering the stakes at the analytical project he has steered at HDFC Securities. The number of customers involved at this analytics project is sheer gargantuan – tapping into 2.5 million customers, their trading preferences, savings patterns, spending behavior, and generating engagement with a platform that multiples money.

Purohit, head for Digital Business & Strategy at the Mumbai headquartered brokerage house offers a pragmatic approach to Analytics in this quick and exclusive conversation with Sairaj Iyer of Sify.com.


Excerpts:


Give us a brief idea about the analytics project at HDFC Securities:

Analytics has a dual role at HDFC Securities. On the acquisition side, it has replaced the manual process of asking customers to open demat accounts. Thanks to analytics we can offer customers a customizable financial goal, plan and track affordability. We also have robo-advisory tools for helping customers track their wealth.

In the post-acquisition stage, analytics is used in engaging with customers on a regular basis. We engage with the customer basis the trading behavior, profile, goals etc. The tool helps us reach out to customers on newer channels such as Whatsapp. We also use the analytics tool to regularly reach out with engaging messages to a user-base of 2.5 million customers primarily banking with HDFC bank. The engagement is various messages and customised offers.

We also have an AI/ML (Artificial Intelligence / Machine Learning) enabled algorithm called Pro-trader to engage with traders.

When you say millions of customers, what is the approach to handling volumes?

I think there is a myth in the direct-proportionality between data and an analytical tool. There is a need to look at relevant data rather than at the ocean or size of the data. Difference between good data and relevant analytical tool is how one can disassociate noise from the system. If you know what you want to achieve, data or the multiplicity will always be of advantage than being a disadvantage. There is a need to understand on how to remove this noise from the system.

Could you elaborate on that a bit?

By removing noise, the system builds better efficiency. Once you know what to expect from the Data, it becomes easier to pick up relevant elements. As a practice, we have been devising the analytics tool to remove noise from the system.

In fact, I don't analyze or mine the data to find the end but rather approach the analytics engine with a problem statement to reach to a solution.

Tell us a bit about the customer engagement?

90% of customers we acquire are first time investors. Starting from teaching them the investment ABCs to hand-holding them through their various goals, there is a continual engagement across various channels.

Surprisingly in our case, we have the highest average incomes on investments from customers who are as young as 12 months old with the system. Customers from other brokerage houses tend to trade with a traders mindset and book heavy losses in the initial few days thereby denting their confidence.

At HDFC Securities, it is the exact opposite. The younger customers are primarily engaged with the cash-segment, once they complete a year with us, they move into other trading products such as derivatives.

How do you handle competition amid increasing robo-advisory services?

Robo advisors can offer inputs based on past-events and algorithms but cannot be a fool-proof system of timing the markets. At HDFC Securities we have a 40-member strong research and analysis team in addition to robo-advisors to offer credible trading acumen and research.

Moreover, our biggest strength is the highly-trusted HDFC network. And, investment for customers coming from this channel cannot be an OTC (Over the Counter) product. There are differences in behavior, goals and other patterns. This is where the combination of research plus robo-advisors makes a difference.

Tell us a bit about the experience of your investment platform that allows customers to invest in global markets?

The response to the Global investment platform in the initial week was extremely overwhelming and we have seen demand pick up. There are no boundaries in the digital world. Names such as Amazon, Uber, Google are quite popular and investors certainly seem keen to invest in their stocks.

The service launched with a US broker and a Fintech provider allows customers to invest in US stocks and ETF (Exchange traded funds) up to their LRS limit of $ 250,000. Being a digital format, it allows customers irrespective of their bank accounts to trade in international stock markets. The customer only needs to open an account for the global investment platform and transfer funds needed to commence trading.

Where do you see the journey of analytics going in the short-term?

I think Analytics at HDFC Securities will go sharper and more incisive. It should help us predict what the customer needs, when the needs arise and the ability to provide that information at the customer's preferred channel and time.