From On-Demand to Instant:The Data-Driven Power behind Payments

Real-time data has become the quintessential experience in contemporary business. An online-only service must operate at the moment or risk losing customers, who now expect round-the-clock access to services.

This urgency was intensified during the pandemic as the shift to online business prompted organisations to invest billions in IT projects to become "digital first."

In the post-pandemic landscape, analysts forecast "aggressive" investment in digital initiatives as businesses enhance their strategies. However, the operational dynamics have evolved.

Merely being online is no longer sufficient. Why? For one, there is heightened competition: 89 percent of organisations are pursuing a digital-first strategy.

Moreover, businesses now have innovative tools to capture sales. These include URL- and QR-code-based payment options embedded in social networks, digital wallets from Apple and Google, infrastructure projects like the European Payments Initiative, and Buy-Now-Pay-Later (BNPL) credit options.

Additionally, customers have become significantly less loyal compared to the pre-pandemic era. Merely retaining customers is no longer adequate. Businesses must develop robust "commerce facilitation" rather than offering a "discrete payment experience."

Initial real-time payment growth has primarily occurred in peer-to-peer settings and online transactions. The next challenges will be in consumer-to-business point-of-sale and billing spaces, which present more straightforward paths to monetisation.

A new wave of real-time commerce is approaching, and riding it successfully will require smarter and more responsive customer engagement.

Banking in the Moment

Some organisations have already started moving in this direction. For instance, BNP Paribas has developed applications capable of making bespoke loan offers to customers at its ATMs, resulting in a significant increase in customer conversions.

Achieving this level of real-time engagement necessitates a comprehensive and constantly updated understanding of the customer. This requires harnessing two sets of customer data: their real-time clicks and streaming data, and their historical data. This information must be blended and analysed using analytics tools at sub-millisecond speeds to deliver actionable, context-based insights that allow businesses to engage customers in the moment and make offers that close transactions.

The good news is that seven in ten organisations believe that, armed with critical customer information, they can make special offers and close deals at the time of engagement. The bad news? Four out of five struggle to unify real-time and historical data to engage with prospects, thereby missing revenue opportunities.

The Challenges

A significant issue is the decentralised and dispersed nature of data. Cloud, social media, and IoT mean data is generated across the IT estate, making real-time data capture and processing challenging. Meanwhile, historical data is stored in customer or inventory databases or in shipping and payment systems on disk-based CRM and ERP systems that are slow and difficult to access.

Additionally, streaming and historical data must be integrated and processed at sub-millisecond performance levels. Integration points between systems can create bottlenecks that hinder analytics and application performance. Moreover, there are computational and security challenges in processing data in highly distributed networks, especially at the edge where customers are but processor resources are scarce.

Platform Thinking for Data

Overcoming these challenges requires a platform-level approach. This involves creating a common and pre-integrated data processing, analytics, and computation environment that breaks through data and system silos to consistently ingest and enrich streaming and historic data while delivering reliable and consistent performance.

What Does a Real-time Platform Look Like?

It has two core attributes. The first is a unified data storage and execution engine for both streaming and historical data. This allows applications to act on data as it is created or captured, rather than processing it offline. Your engine should enable data streams and threads to execute concurrently and seamlessly distribute work for performance, scalability, and responsiveness. Integration at this infrastructure layer frees IT teams from building and maintaining complex integrations with their inherent performance bottlenecks.

The second core feature is an in-memory computing architecture. In-memory data stores allow data to be accessed and processed in RAM, providing the fastest possible speeds and eliminating the wait times associated with retrieving data from slower media. This is crucial for real-time analytics, applications, and payments. An advanced in-memory data store can cluster nodes and pools of memory to provide local computational power and high application performance, as well as a caching layer for microservices. In-memory systems thus offer the performance required by real-time analytics, applications, and payments.

In-memory systems also enhance data security. Because data is stored in memory rather than on disk, payment processors do not need to store sensitive information, such as payment card details, on persistent storage, reducing potential avenues of attack for hackers and facilitating easier compliance with privacy regulations.

With the rise of digitalisation comes a new wave of real-time commercial opportunities. By leveraging a unified platform of data analytics and computation, businesses can deliver the intelligence needed at the speed required to succeed.