Talks Tech #39: Raka Khushu, Vice President of Analytics at EXL
Written by Raka Khushu
Women Who Code Talks Tech 39 | Spotify – iTunes – Google – YouTube – Text
Raka Khushu, Vice President of Analytics at EXL, discusses the importance of analytics, how to use them to improve processes, the best way to interpret raw data, practical applications, why its important, and how to practically implement data in a company’s strategy.
Data analytics is collecting, organizing, analyzing, and transforming raw data into comprehensive information. That transformation aims to increase a business’s performance or efficiency by bringing about insights and value and aligning objectives to business goals in the long run.
This requires information technology because you need access to parse through various kinds of data to your business objective, a knowledge of statistics because you end up applying many sophisticated or simplistic techniques to the data, and business acumen. Ensure that you’re able to connect the dots and drive organizational goals forward. In an ever-evolving analytics industry, you also need the ability to communicate this information and derive insights from it to understand which direction to move.
At a high level, analytics can be categorized into four areas depending on the objective. The first is descriptive analytics, which describes what happened in the analyzed period. For example, the number of new crest card users has increased. Whatever allows you to describe the data in question would be labeled as descriptive analytics.
Then diagnostic analytics help you to understand why the data is showing that. How data is described by diagnostic analytics tells you why something happened the way it did and allows you to find the root cause for troubleshooting or subsequent planning purposes.
The next step is predictive analytics, which is all about forecasting what might happen in the future using statistical techniques, data mining, business assumptions, market conditions, and a very important historical trend. An example of this would be credit card companies that want to know how many new customers will be able to pay their bills and how many will default.
Last but not least is prescriptive analytics. Once you’ve done descriptive, diagnostic and predictive, prescriptive analytics is all about recommending future courses of action to be taken in case certain situations arise and the potential implications that could come from them.
In our credit card example, if you predict that sales will go down, how do you deal with it? If you expect a positive outcome, how do you handle that, and what does that mean for the business? This allows you to take strategies to mitigate loss or optimize gains.
Companies are increasingly investing in data analytics. It allows them to measure, monitor, and plan strategic aspects of the business. They let you obtain deeper insights into the key metrics associated with the company and a mechanism to link and label them to the business objectives. Having access to those deeper insights is the key to making faster decisions. Knowing the key drivers of a situation and prescriptive analytics lets companies make much more nimble decisions because they have access to all of the supporting facts.
It also provides streamlined operations because it allows businesses and companies to manage their logistics and operations effectively by identifying, planning for, and resolving bottlenecks and making sure that they know what kind of cost they will incur. Being able to prepare for bottlenecks, identify redundancies, and figure out ways to remove them allows you to reduce operational expenses.
Perhaps the most important benefit of data analytics is issue management. Businesses can identify, resolve, and document anomalies and issues and have appropriate procedures for handling them, allowing them to be better at operational excellence proactively, identify and resolve any customer-facing gap system issues that they may come across, and bridge them before they result in a full-blown issue, or cause any regulatory harm.
Next, I’d like to discuss what analytics looks like in practice. Banks have been known to use customers’ credit history and spending data to determine who qualifies for a loan and what the loan terms are. Retailers have also used customer data and segmentation to improve retention and loyalty.
Credit card companies also use spending data and the demographics of customers to detect fraud. In an industry where fraud is pervasive and fraudsters are increasingly becoming smarter and smarter and using more and more sophisticated tools and techniques to avoid detection, it becomes essential for banks and credit card companies to stay on top it.
How customers are spending, how much they’re spending, and how far away they are from where they live geographically are important factors in determining whether or not there is a fraudulent transaction happening on their account. That gives companies not just a safeguard but also protects the interests of their customers, which is a very important factor in customer loyalty and retention.
Analytics is more rampant than you think it is. Anytime you’re collecting data, think of it as applying some form of analytics.
Next, I’d like to talk about the credit lifecycle. The credit issuer starts by targeting a group they want to give a credit card. This includes the product they want to offer, the terms, and how they should be targeted. This is based on a customer’s creditworthiness and what kind of product they might be inclined to use. Would they respond more to a premium, rewards, or mass market product?
Once someone has been targeted and they decide to apply for a credit card, the next step is underwriting, which is all about asserting whether the company can approve that application. That uses expected profitability over the lifetime of the relationship, meaning the revenue a potential customer would have over time starting with the formation of a new relationship with the credit issuer.
The next step in the journey is limit assignment and pricing. A lot of people will be familiar with this. Most of us have credit cards, which all come with a particular line or credit limit determined by creditworthiness. This considers what credit limit would best encourage the customer to spend higher dollar amounts within safe criteria. This has to be a balance between how much they can spend, how much they can afford to pay back, and the potential for fraudulent use.
After that the credit issuer has to consider contact strategy. This is about when a customer should be contacted, how they should be contacted, and what information should be used in that engagement. This can include things like what is on the monthly statement, and what should be included on it based on their behaviors. It can also have any concerns about fraudulent activity. Some of those communications can also be optional. Customers can choose to get alerts around spend activity and balances.
We then have to consider how to manage customers throughout the relationship effectively. This includes monitoring the health of their portfolio so that you’re able to action important items accordingly. If customers are high value, meaning that they’re low risk and spending high, then should you focus on positive treatments for them including credit limit increases or promotional offers.
These controls allow for proactive mitigation, resolution, and bridging of system issues or errors that can lead to customer harm or tarnishing the company’s brand image. Analytics are involved in all of those areas fundamentally.
Finally, I want to share why we must constantly stay on top of things and innovate even within the analytics industry. Companies need to remain nimble regardless of the industry. Innovation and transformation have become buzzwords these days, and they’re essential because companies need to be robust, able to adjust for disruption and have practices in place to ensure continuity.
The industry is changing. There’s a lot of fast-paced innovation, so this is by no means the end, and there’s more coming. It’s a perfect industry to be in, with a lot of potential for leadership.
***************
Guest: Raka Khushu, Vice President of Analytics at EXL
Producer: JL Lewitin, Senior Producer, Press and Digital Content, Women Who Code