a

Lorem ipsum dolor sit amet, consectetur adicing elit ut ullamcorper. leo, eget euismod orci. Cum sociis natoque penati bus et magnis dis.Proin gravida nibh vel velit auctor aliquet. Leo, eget euismod orci. Cum sociis natoque penati bus et magnis dis.Proin gravida nibh vel velit auctor aliquet.

  /  Project   /  Blog: How Analytics can be used to mitigate risks in Financial Service organizations

Blog: How Analytics can be used to mitigate risks in Financial Service organizations


Just the other day I was in conversation with one of my tennis buddies after an invigoration set of three matches. He was facing a very commonplace yet serious issue at his workplace — a financial services organization which provides efficient assistance with procuring various financial products, loans, credit cards etc. A separate customer service division would provide assistance to users — helping them with any issues and to help resolve complaints if any.

However this team was unable to effectively track customer complaints over a period of time — a lot of complaints were being closed without an explanation and even worse some complaints had to be closed via monetary settlements! It was very difficult for them to analyse the numbers on massive excel sheets with multiple formulas. Also spreadsheets aren’t designed to store historical data, so often, in an attempt to keep the size of them manageable, they are “updated” and companies lose their historical data. Another drawback is that not everyone can identify the errors in a complex formula which leads taking wrong actions based on data misinterpretation

It was at this time that we demonstrated to his team how ZEPTO could be used to effectively avoid such situations. Zepto is a self-serviceable analytics tool for small/medium businesses in the financial services sector. It’s AI driven Insight generator aids decision making for a SMB.

Let me take you through a deep dive of this use case. The very first thing that we did was to analyse the complaint data which was being monitored on an excel sheet. This excel sheet was uploaded to ZEPTO and we determined that there were 157, 000 complaints overall. Great! Now let’s have a look at the 5 main issues which were not being addressed effectively by the company:

1. Are the complaints responded to within the stipulated time?

2. How frequently are complaints being received?

3. How effectively have the complaints been closed?

4. How many complaints have ended up with a monetary settlement to get them closed?

5. What are the divisions of the company that are involved in the most monetary settlements?

Let us deep dive into how each of these problems were addressed.

1. Are the complaints responded to within the stipulated time?

Using Zepto we needed to visually analyse if complaints are being actioned upon in a timely fashion. By using simple drag and drop features we came up with a pie chart to see that 98% of the complaints have been responded timely. But is it consistent over months?

The chart type can be easily changed to ‘hundred percent bars’ with the date field also added to the second cage. To see the distribution, the ‘timely’ field was added to the third cage. Now to analyse this across different months, we simply selected the desired granularity from the auto- categorized date aggregations. The results were that the response rates were steady except for the last two months which fell below 98 percent.

2. How frequently are complaints being received?

We generate a simple chart to show the flow of complaints over the months of the year. We noticed a pattern where complaints are high in the mid of the week especially on Tuesdays.

Another important feature of ZEPTO which we displayed was to get a forecast of the number of complaints for the upcoming days. Unlike in other products which show a linear, polynomial or exponential trend lines, ZEPTO’s AI model can understand the seasonal pattern and mimic the same in the forecasts which makes it realistic.

3. How effectively have the complaints been closed?

While analyzing the data we wanted to check which medium is effective to receive complaints. It was evident that the web is dominantly effective in receiving complaints

To determine how effectively the complaints were closed, a pie chart was generated. Pretty impressive as more than 81 percent of the complaints were closed with an explanation.

4. How many complaints have ended up with a monetary settlement to get them closed?

To determine how many complaints ended with monetary settlements a basic filter was applied to the data and we were able to present the numbers. Whoa! More than 7000 complaints ended up with monetary settlements.

5. What are the divisions of the company that are involved in the most monetary settlements?

Checking account and General Purpose Credit Cards divisions has the most of the issues that resulted in monetary settlements.

We further analysed the divisions that resulted in a monetary settlement on a month by month comparison for the last two months. On applying filters on the date field in multiple ways we came to the conclusion that Credit & Pre-paid cards division has seen a huge jump in the number of complaints.

In this manner financial organization was able to use ZEPTO to streamline processes and mitigate risks . This has helped them to get insights (within a couple of clicks) that could be easily actioned upon!

For further details on ZEPTO refer to : https://www.zepto.io/

Source: Artificial Intelligence on Medium

(Visited 4 times, 1 visits today)
Post a Comment

Newsletter