What is the analytics roadmap that you have set for your organisation?
Our data analytics roadmap aligns with the overall corporate strategy and business objectives which can be broadly classified as – top line growth, cost reduction and risk and compliance management. We are leveraging data analytics in our day to day functions such as acquisition, activation and continuous meaningful engagement with our clients. This will help us in keeping up with the changing consumption patterns of the customers as well as addressing the business dynamicity of the capital markets. For example, the product recommendation to a short term investor vis-a-vis a long term customer will be different. I will not be able to engage a short term customer with an SIP and likewise an options trading product will not fit a long term investor.
We run a lot of propensity models and identify the best investment mix for the customer on the basis of his investment philosophy and what we can offer. HDFC Securities is an Amazon of financial services. There are a number of investment options that we can identify for the customers and offer him the best fit.
What is the kind of infrastructure built for analytics? Data integrity is a primary requirement for analytics. What steps have you taken for having accurate data?
We maintain the data integrity at the core platform level and have built various intelligence layers over it with the use of BI software. A layer of analytics is built over both the demat and materialised investment platform. A complementary partnership structure has been created with the fintech companies in order to get other associated customer data. Post which, various analytical models are run to create customer persona and then do customer segmentation.
The benefits from these initiatives have been immense. Two years ago, the percentage of transactions on the digital mode were 60 per cent of the overall transactions using other channels. Currently, it has risen to 83 per cent. Additionally, the overall revenue has seen a spurt of 40 per cent.
From an institutional brokerage perspective, we do algo trading for four major clients and have a dedicated analytical platform for providing market insights to the customers.
What is your team composition as far as the analytics team is concerned and how are you getting the best talent from the market since there is a paucity of talent in the analytics space in India?
We have a team of four-five full time data analysts with us, drawing inferences from various customer data sets – be it behavioural, consumption or in some cases – predictive. Apart from this, we have interns from various data science institutes, both national and international. Data gathering and indexing is automated by using the BI tool sitting on top of the analytics layer. Sixty percent of the team is engaged in creating propensity models. The rest 40 per cent works with the stakeholders to provide them insights and inputs.
The team works in close loop with the relevant teams from HDFC Bank because the customers are predominantly from the bank. The propensity models are refreshed and cross checked with the teams from the bank.
As head of analytics, I decide the right propensity model to be applied for a particular business problem. The key to bringing efficacy in using analytics is by correlating data with understanding consumer behaviour. The business teams come up with their requirements of the products to be sold and the analytics team suggests on which products to be sold to which set of customers.
Please highlight the initiatives taken in the space of analytics in your organisation
Few Highlights of initiative are mentioned below:
- Using BI software layer on top of data and provide access to key stakeholders in terms of output visualisation, reports and dashboards
- Using various analytics model like predictive modelling, look alike model, KPI projections, churning prediction, propensity modelling, etc.
The company is in discussions with various third party players for understanding of use cases and undergoing various POCs. Using the fallout assessment model, HDFC Securities is doing a PoC with customers who were of high relationship value but have reduced the engagement in the last few years. They are showing early signs of attrition to the competition. This PoC has been running since the last month and a half and the company has been able to win back 14 per cent of the targeted customers.
In another PoC being run for over six months now, non profitable acquisition has been reduced by 30 per cent. This is as a part of the efforts to finetune the customer acquisition process. Using analytics, the conversion ratio in the space of marketing has doubled in the last one year.