For more than 25 years, Naranja X has been the primary credit card issuer in Argentina, with its ubiquitous orange cards in the hands of millions of Argentinians. But the leaders of the 37-year-old company saw the writing on the wall: Customers began to prefer to get all their financial management products and services digitally and securely, from the same vendor.
For Naranja X, getting to that point meant shifting its business model to one of a Fintech company and using technology to automate the delivery of financial services. The company wanted to expand its online ecosystem of products and services to include transfers, collection offerings, loans, insurance, e-commerce, payment services, travel, and promotions.
It’s a smart move for well-positioned companies. The global Fintech market is projected to grow at a compound annual rate of about 25% annually through 2027, according to Market Data Forecast. That growth will be spurred by the digitalization of financial services, the demand for contactless payments, and Fintech’s potential to attract more customers.
But switching its business model to Fintech required more than the will to do it. Naranja X had to to take a hard look at its data and how it could squeeze more value from it.
“We needed to become a data-driven company, and that meant that everyone in the company needed not only to have access to data but to understand the data,” said Lucia Arando, the company’s data governance manager.
Naranja X also needed new processes to guarantee data security, which involved restricting certain data to the right people for the right purpose. And that data had to be fully controlled and accountable, because Argentina’s Central Bank puts many more requirements on Fintechs than it does credit card companies.
Data Management Challenges
Preparation for becoming a Fintech company started about two years before the switch itself. During that time, Naranja X worked to better organize and manage its data.
First and foremost, that meant being able to quickly search for any data asset in the company. Naranja X traditionally relied on Excel spreadsheets, using one for each data domain -- for example, one Excel file contained only client data, one only supplier data, and so on. While searching for data in a specific domain wasn’t difficult, searching for specific data across more than one domain was virtually impossible.
The new data management system also had to automatically track any additions, deletions, or other changes to existing data; log user activity; and facilitate collaboration around data.
An added obstacle was the fact that data stewards -- the main owners of data in each domain -- didn’t really understand why they needed to keep on top of their data so closely. “Our data governance program is based on stewardship, but because we didn’t have anything tangible, our data stewards couldn’t understand why they should do what we asked them to do,” Arando said. “They didn’t see the value.”
What Naranja X needed was an automated method to create a comprehensive data catalog, provide secure data access, and ensure full data governance. To buy time while evaluating potential products, Arando’s team chose to temporarily shift to Microsoft Power BI. The team uploaded its existing Excel spreadsheets, each for a different data domain, into Power BI. This enabled users to see all metadata in one large table with a search box and filters.
While Power BI solved some problems, it was far from the optimum solution. Not only did it only support front-office metadata -- the back office was still fully manual -- but data stewards didn’t fully accept it.
It was during this time that Naranja X’s data management team learned a valuable lesson. “We needed to ask our users what they wanted to have that they didn’t have and what they didn’t like about the current processes,” Arando said.
Implementing OvalEdge Data Governance Tool
Naranja X’s first step for its transition was to find a way to automate the collection and management of technical metadata -- the technical properties of data, including location, credentials, source, schemas, and attributes.
The technical metadata tended to become outdated quickly, making it a significant data governance challenge. Relying on data stewards to inform the data governance team when things changed was hard to enforce and not consistently reliable, Arando said.
The team decided to standardize on OvalEdge, a data catalog and governance tool that automates the process of locating and displaying technical metadata. It was a good start, but the team also needed a way to automate collection and access of business metadata. OvalEdge could handle the business metadata, but only if Naranja X data stewards agreed to consistently enter the data into the OvalEdge platform.
Given human nature, that was a big ask, but it was important enough that the data governance team came up with a plan. Using Power BI, the team created a game with incentives for adding the data to OvalEdge. The game was a simulation of Formula One racing, where drivers are data stewards and their activity in OvalEdge can earn them points. Every quarter, the company now announces the top three winners, who receive bonuses in their Naranja X app.
Those gamification efforts clearly paid off. Usage is up significantly. In addition, activity on the company Slack channel for data questions and answers has decreased significantly, with users becoming familiar with OvalEdge’s self-service options.
Today, most technical and business metadata is kept in OvalEdge, along with advanced analytics models the company has built, organizational KPIs, queries, and reports. The tool is integrated with Naranja X’s AWS-based data lake, as well as with an onsite Oracle data warehouse that will eventually be phased out (all data will move to the data lake). The data governance team has set alarms in the system to trigger data governance actions when a change needs their attention, such as when a user deletes or adds a table.
Internal users can easily search for whatever they need. For example, if a user want to search for information related to all active clients, a simple search will find every dataset related to that term. Data analysts and data scientists also can easily access statistical data.
“Having all of the data in one place where any data scientist can go to understand key data assets is important, especially since data scientists are difficult to retain,” Arando said. “When a new data scientist comes to the team, the first thing they want to do now is see what we have in OvalEdge.”
The new system has also improved data security, an important requirement for data governance. For example, the organization needed a method to quickly identify the location of sensitive data and make sure that it is secure in terms of obfuscation. OvalEdge takes care of that by obfuscating strings with an algorithm and then integrating it into Snowflake, the exposure layer, where end users consume data.
Now that many of the data management and governance challenges are under control, the next step is to integrate back-office sources and functions into OvalEdge. Naranja X will also identify and classify all confidential data and incorporate transactional data. The goal, Arando said, is to create end-to-end data lineage.