Laying the Groundwork for the Next Wave of Growth
- By Paul Mah
- January 10, 2023
Around the world, growing geopolitical tensions, soaring inflation, and softening markets are coming together in a perfect storm of uncertainty. As an industry, the banking, financial services, and insurance (BFSI) sector must move quickly and decisively to navigate this challenging business environment.
At a face-to-face event hosted as part of the Informatica World Tour in Singapore, BFSI executives and leaders gathered to share their experiences and to discuss how data coupled with the right tools and culture can lay the groundwork for spectacular growth despite the bumpy season ahead.
Timely access to data
Data has always played a pivotal role in business, though how it is stored has changed even as its use has expanded dramatically over the years, says Gary Ho, the chief technology officer for Hong Kong & Macau at AXA.
Speaking as a panelist during the panel discussion segment, Ho said: “In past years, a lot of the data came from core systems such as our AS/400 and policy admin system. Now it's our data lake and big data platforms.”
Accessing data was far less intuitive then, and requests for data could take hours or days. Today, many of the usage scenarios around automation, personalization, or creating new customer-centric products require timely access to data, he said, which makes real-time data access a business advantage.
The use of data has certainly evolved. Speaking on this topic, an executive from a local bank spoke about how his organization built a “ringfenced” sandbox to support data discovery. The controlled environment made it possible to give employees timely access to data – and the opportunity to build dashboards, analytics, and even machine learning models.
There were limits to this approach, however. “The challenge came with machine learning models; how do you operationalize them quickly? Users struggle with that. But having users create their own dashboards and reports is quite feasible today,” he said.
Understanding the other side
It is also important that data teams understand what the business is trying to do, says Peter Ku, the vice president and chief industry strategist for Financial Services at Informatica. He said: “Otherwise, these investments [in tools to manage data] will not be funded in today's climate.”
And while technology groups should seek to understand the business priorities, it helps if business users also understand the challenges that technology teams face. According to Ku, this can be achieved by creating opportunities for employees to better appreciate the work of their colleagues.
Ku recounted his experience at a previous financial organization he worked at: “We started this rotational program where we bring the data teams and rotate them to the business side of the organization. They saw what their business stakeholder teams were doing with the data they provided. At the same time, we had business groups sit in on technology conversations, to better understand the systems that were working behind the scenes.”
“They knew that the car must go from point A to point B. But they don't know what's happening under the hood. And after witnessing the complexity, they are less likely to take future conversations for granted. I encourage organizations to give business users more exposure to the technology conversation,” he summed up.
The importance of data quality
Ku also shared an anecdote of a conversation he once had with a CEO who insisted that his organization had all the data dashboards and reports he needed. After getting permission to speak to the employees tasked with generating his reports, Ku sat down and spoke to about 20 of them.
“And I found out that these employees are all financial analysts, people that are responsible for their cash flow statements or balance sheet reporting. When I asked about the percentage of data that was either incorrect, invalid, or outdated, the figure was 90%,” said Ku.
It turned out that the data sources were just not good enough, and the analyst had to fix the numerous errors by exporting data from the ERP into a spreadsheet to manually correct them. While all errors were fixed prior to the reports being generated and sent to the CEO, the manpower cost was huge with between 50% and 75% of their time spent fixing data issues.
“We had to educate them that the reason these data problems exist was due to a lack of proper governance or data management processes upstream. I still remember the puzzled look on their face when I asked them to imagine a world where the data was complete, correct, accurate, and fully defined – they thought fixing data problems was their job.”
With the right processes and systems in place, analysts could conceivably move on to high-value analyses that substantially benefit the organization. Of course, this does entail a cultural change in how organizations manage their data.
The time to get started is now
The other panelist, Balaji Narayanamurthy, the president and head of Business Intelligence at Axis Bank, shared his insights on data management. While dashboards and machine learning models look appealing, Narayanamurthy’s advice is to start early on areas like data governance.
“Keep data governance, data privacy, and all the other foundational stuff right up front, because they just become that much more complex to change at a later stage,” he suggested.
Finally, don’t attempt to boil the ocean, advised another executive: “You need to be very tactical about which system to start on. You cannot focus on every single system because the cost would be astronomical.”
“With [a solution] like Informatica as tooling, we can then expedite [data initiatives] by taking away manual processes. Start on something small and make major headways around a specific project with well-defined outcomes. After that, you can use that as a success story to get more buy-in.”
Paul Mah is the editor of DSAITrends. A former system administrator, programmer, and IT lecturer, he enjoys writing both code and prose. You can reach him at [email protected].
Image credit: iStockphoto/RomoloTavani
Paul Mah
Paul Mah is the editor of DSAITrends, where he report on the latest developments in data science and AI. A former system administrator, programmer, and IT lecturer, he enjoys writing both code and prose.