Skip to main content

The keys to next-generation data management for financial services

(Image credit: Image Credit: Pitney Bowes Software)

According to the latest statistics 2.5 quintillion bytes of data are created each day, and that number is set to rise with the growth of the Internet of Things (IoT). This is because today, data exists everywhere, across every vertical.  In financial services (FS), capturing and leveraging massive volumes of data – including  customer information, financial transactions, product and service purchase histories, marketing campaigns, service inquiries, market feeds, social media streams, IoT streams, software logs, text messages, and new sources such as images, audio and video – enables companies to capitalise on new data-driven business opportunities.

However, connecting data from disparate sources poses a huge management challenge for FS companies. Not only do they need to be able to derive insights from data stored on-premise, in the cloud, and in hybrid environments. They also need to ensure they work with data they can trust, with a transparency that helps them verify data origin, quality, and traceability.

To satisfy these needs, companies need a modern data management strategy that addresses both enterprise data – relating to customer information, contractual agreements, and financial transactions, among others – and Big Data – encompassing semi-structured or unstructured data, such as emails, social media streams, as well as image, audio, and video files. Current data management practices often fail to create a link between enterprise data and Big Data. This makes it difficult to operationalise data science and derive the valuable benefits of data-driven analytics. As a result, companies may struggle to search massive haystacks of data to find the hidden needles of actionable insights. This missing link also prevents companies from delivering data-driven innovations, which are a core ingredient of digital transformation.

To reduce complexity, companies need to combine their data into a single universe. Universal data helps firms enhance visibility, delivering insights that can improve efficiency, automation, and growth. By converting data into insights, organisations can become intelligent enterprises.

For many FS companies, however, a single data universe is still an aspiration. More often, data resides in multiple siloed environments. Because data is not meaningfully connected across these silos, it has become less accessible – compromising insight into customers, partners, products, sales channels, and financial performance.

Worse yet, data silos often are reinforced by organisational silos. For example, the group managing Hadoop data lakes are not the same people who manage cloud storage. And too often, teams use different tools and rarely interact with one another.

To overcome the challenge of multiple data silos, FS companies tend to build large enterprise data warehouses, which often can no longer keep up with the analytics needs of the business.

These challenges are further complicated by the increasing number of data consumption endpoints, the business processes and analytics solutions that require real-time data access to support decision-making.

Modern data management

FS companies can effectively manage data by implementing the use of tools and methodologies such as machine learning (ML) and predictive analytics. They key is to leverage existing assets, connect data across the whole technology landscape, and simplify data analysis by reducing data redundancies.

With these objectives in mind, there are three main elements companies must consider when creating a next-generation data management approach:

  • A unified logical FS data model supports data consistency and simple access from analytics applications. It also minimises data replication and reconciliation efforts. Modelling the business view of data helps companies take ownership of their data because they are not required to understand the physical implementation on the database level or the complexity of multiple physical data silos.
  • A modern data management platform, built on trusted, connected data. This requires organisations to collect and integrate data in a unified data landscape based on the standardised logical data model.
  • A data hub, which can help companies gain a holistic view of data assets, manage data across the full IT landscape, and integrate data into a unified view. By building the platform around a data hub, they can increase transparency of and access to all data assets, which increases agility and the speed of innovation.

While, each of these data management features is well-known, most companies implement them through multiple tools and technologies, based on the siloed databases deployed across their IT landscapes. Without a unified approach, traditional data management is often complex and slow.

Using ML to drive growth

Once companies create a solid data management foundation they can begin implementing machine learning algorithms to support automated decision-making and data-driven process optimisation – helping them generate insights that create better customer experiences, improve operational efficiency, and drive sales.

ML can help FS firms to deliver personalised services based on customer profiles, using data on customer satisfaction, preferences, buying history, demographics, and behaviour to better understand their needs. These insights can help companies tailor products and services and deliver highly targeted, personalised offers that improve customer satisfaction and retention.

When it comes to protecting the business, among others, ML algorithms help provide early warning predictions using liability analysis to identify potential exposures prior to default. They also predict risk of loan delinquency and recommend proactive maintenance strategies by segmenting delinquent borrowers and identifying self-cure customers. With this insight, banks can better tailor collection strategies and improve on-time payment rates.

Just as importantly, automation plays a central role in data security – and in the business opportunities it presents. Most if not all FS companies handle and store sensitive data, and as such must comply with data security laws such as the EU’s General Data Protection Regulation. Far from being a tick the box exercise, utilising automated security technologies has enabled FS companies to go beyond mere compliance and gain competitive advantage by creating new business models centred on data security.  For example, some banks are using their data management platforms to offer a “digital vault” where customers can store and share sensitive digital assets, much like they used to do with physical assets. Data security is also the key to enriching data and strengthening customer profiles over time – as customers’ willingness to share additional personal information is also be based on their level of trust.

Last but not least, ML streamlines processes, from automated trading and real-time credit decisions, to customer complaint management and inquiry responses.

As a result, as FS leaders increasingly realise that more trusted, connected, and intelligent data contributes to digital transformation, they are viewing data not as a cost driver, but as an essential asset – one that requires certain investments to unlock its true value. Building a modern data management platform – one that enables massive “haystacks” of data to be automatically analysed for hidden “needles” of actionable insight – is an investment that will pay off handsomely for tomorrow’s digital leaders.

Roland Bloesch, VP, Global Head of Regulated Industries, SAP Customer Experience