The benefits of data streaming are numerous—hundreds of organizations have reported increased profitability, better business responsiveness, and multiplying ROI. But understanding just how these organizations achieved these results—and the practical tips and steps to get there—is where things can be a little trickier.
At this year’s Data in Motion tour stop in Singapore, we convened four experts across retail, financial services, agro-industrial, and management consulting to discuss the strategies they’ve adopted to drive efficiency and innovation through superior information and customer experiences. Stuart Ward, Confluent’s Financial Services Industry Lead APAC, asked these leaders about the importance of data in their organizations and how they’ve been able to deliver it more efficiently through data streaming, data mesh, and more; using data to create excellent customer experiences; best practices for modernizing legacy applications; and more.
Delivering real-time customer experiences
Making real-time data valuable to banking customers means starting with the basics: the data has to be accurate and secure, said Stuart Loxton, Data Lead at Trust Bank. From there, Trust Bank—Singapore’s first digitally native bank—is able to enrich the data, such as through transaction categorization and cleaning up merchant names, adding another level of value for customers.
For the Central Group, delivering value means not wasting customers’ time, said Clovis Warlop, Head of R&D. “We make sure that when you come on our websites, you find the product you want and you find it quickly,” he said. “We make sure that every time you come back we know what you want.”
To make that happen, the Central Group—which serves shoppers in Thailand—has a system of attribution that is able to track where every customer is coming from, along with several different data sources that build a profile of the customer and their interests. Even things that don’t seem relevant end up being relevant most of the time, he said.
Then, Warlop’s team is able to enrich the data by identifying buying trends: what gets bought together or not, or what is bought often. All those insights are combined to allow the Central Group the ability to see a customer’s “propensity to purchase.”
But the need for excellent customer experience isn’t just in retail—even for government agencies, citizens expect proactive services, especially after the pandemic, said Johnson Lam, Managing Director at Accenture. “They expect that the data is actually already available even before new services launch.”
The importance of modernizing applications not just for customers, but also employees
Singapore has a talent crunch and an aging population, Lam said, so efficiency has been a major focus to help civil servants do their work faster. Organizations are shifting toward a system where the offices only interact with citizens at a point in time when they need to, he said, so that agencies can reduce overhead and the cost of running their offices.
In Thailand, delivering food to citizens requires an extremely long supply chain with many players, from the feed business to slotting to food processing, said Chalerm Duangyeewa, IT Consultant at AXONS/CPF. The challenge, he said, is “how can [all these businesses] use the data, and how can we organize it in a timely fashion and [make sure it] is correct or it's immutable, meaning we can trust it everywhere.” To reach that goal, Duangyeewa’s team turned to Kafka, shifting from a point-to-point interface to a synchronized pub/sub integration.
Why data mesh empowers everyone in an organization
“The thing that we really like about it is that every team becomes responsible for their data,” says Warlop, from the Central Group. “The one who creates the data is also responsible for making sure the data gets on the platform, making sure the data's in good shape.” A decentralized system like data mesh allows the Central Group to iterate and innovate quickly, so smaller teams can hop on projects independently without requiring the whole team.
That’s beneficial for both customers and staff. Recently, someone on Warlop’s team was able to build a new service for customers called “Shop the Look,” which takes a picture of someone’s outfit and then suggests what pieces to buy.
At Trust Bank, data mesh helps build automation and efficiency, Loxton said. “Rather than relying on a series of signoffs that aren't really mapping to your organization, you get a clear idea of the ownership of the data,” he said. “We know where the data is because we have a central repository of all our schemas tagged specifically by bounded context.”
Advice for using data streaming in your organization
Loxton said to take one step back before jumping right to streaming. “When you create software, what you're really trying to do is model business interactions,” he said. Upon thinking about what you actually want to do, streaming is often a solution that comes out of that, he said. Rather than think of integrations or technology, think about the key business interactions you want to have and how you can use events to reach those outcomes.
Lam said that governments and other organizations with legacy systems also need to be careful about modernizing to make sure they’re modernizing the application to help the organization work—and not the other way around. He later added that organizations need to identify who are really the decision makers, who gets the call for a particular domain. “You don't want to be a decentralized system but end up having 50 people to ask in order to make a single decision,” he said.
Watch the full panel session, and catch up on other topics we covered on the Data Motion Tour. For more insights and information on how business leaders are seeing real-world benefits by adopting real-time data streams, check out our 2023 Data Streaming Report.