Being data-driven is firmly on the radar of most financial institutions. They recognize the value and power of data and use it to make better-informed decisions, streamline operations, and improve the customer experience. In short, data is a competitive advantage – when harnessed optimally.
Indeed, according to a 2022 survey by NewVantage Partners, 92% of organizations gained measurable results from investing in data-driven technologies – a marked increase from 48.4% in 2017 and 70.3% in 2020. However, to fully empower themselves and generate the most value from data, banks should consider their strategy – notably, whether a decentralized approach, notably data mesh, could benefit them.
Discussed at last year’s Summit with Lisa A. Schiborr, Senior Consultant at Sopra Steria, and Blaise Ngonmang, Data & AI Leader for Financial Services at Sopra Steria, we continue the conversation below.
Transforming into a data-driven enterprise
Banks already leverage data, but there’s room for improvement in many cases. According to Lisa A. Schiborr, there are “hidden data opportunities” and “little gems with the potential to directly impact the bottom line”.
To really capture the value of data, banks should take these two factors into account: technology and culture. A data-driven enterprise goes beyond collecting and evaluating information; it also requires a mindset shift – a culture where data is accessible and plays a vital role throughout the entire bank.
Achieving that involves five pillars that revolve around business benefits and the scalability of technology and culture.
- Single source of truth – centralized platform architecture.
- Data-as-a-self-service.
- Data-as-a-product (DaaP).
- Promotional governance.
- Data democracy – the ultimate decentralized goal.
For pillars one and two, technical transformation is a priority. Each bank is on its own journey, but ones that don’t move beyond pillar one – a centralized architecture – may experience several issues, including the following.
- Threat to data integrity – it’s impossible for central teams to have complete knowledge of all the bank’s data.
- Waste of resources – each use case re-transforms and re-observes the same data.
- Difficulties with cooperation – teams building use cases don’t interact with each other.
- Central data team is a bottleneck – they can’t oversee each domain team effectively or answer questions from the C-suite quickly enough.
As banks move into pillar three and beyond, a cultural shift also comes into play and is increasingly important. By the time they reach data democracy, information is widely accessible through the appropriate technical tools, and neither a profound knowledge of data nor analytics expertise is required by end users.
Alongside that, leaders are motivated to help change the mindset of their teams and the rest of the bank. To that end, they encourage data literacy and data-driven decision-making and ensure data stays accessible and trusted.
What is data mesh?
As banks move along the axis from a centralized to a decentralized approach and strive for a cultural transformation, data mesh is a consideration. Coined by prominent technologist Zhamak Dehghani in 2019, it’s based on four pillars that work together and bundle well-known concepts.
- Domain ownership (decentralization): Knowledgeable domain teams are responsible for their data from an analytical and operational perspective, rather than the central team.
- DaaP (product thinking): Standalone data sets are created and maintained with end users in mind and an emphasis on quality, usability, and satisfaction.
- Self-serve data infrastructure (platform thinking): The whole bank can seamlessly access and process data using domain-agnostic tools and systems provided by a dedicated data platform team.
- Federated governance (system thinking): Standardization across the whole data mesh achieves interoperability of data products and an ecosystem that adheres to the bank’s rules and industry regulations.
When is data mesh the best approach for banks?
Moving to a data mesh approach is a transformation journey that builds a new relationship between the business and the technology. It comes with a myriad of benefits, from increased flexibility, regulatory compliance, and reduced time to market to improved data quality, ease of system thinking, and cost savings.
Moreover, banks don’t need to implement all four pillars simultaneously to start reaping rewards – the strategy can be piecemeal. For banks considering data mesh, Lisa A. Schiborr believes there are questions they should ask themselves around five key dimensions.
- Organizational complexity: Is the bank scaled enough in terms of its data? Does it have pre-existing centralized data management solutions that may inhibit the success of a data mesh approach?
- Data strategy: Is there a plan in place, and how broadly does it apply across the bank? Is there room for change and improvement (and a commitment)? Does the bank want to generate value from data at scale?
- Leadership support: Is the bank ready for cultural change? Are the leaders equipped to deal with resistance? Is there C-level support for a mindset shift?
- Domain-oriented: Is the bank modern and digital-forward? Is it designed according to business domains?
- Data technology at the core: Are artificial intelligence and machine learning at the heart of the bank? Is the institution determined to take advantage of data in the long run?
Investment on returns using a data mesh approach
Several banks have successfully implemented a data mesh model – we briefly explore two examples below.
JPMorgan Chase (JPMC)
JPMorgan Chase identified a paradox: “Data that is permitted to be freely shareable across the enterprise has the potential to add tremendous value for stakeholders, but the more freely shareable the data is, the greater the possible risk to the organization.” To solve that and unlock the value of their data, JPMC uses a data mesh architecture to align their data technology to their data products.
As a result, they “enable data sharing across the enterprise while giving data owners the control and visibility they need to manage their data effectively”.
ABN AMRO
Meanwhile, ABN AMRO uses data mesh alongside a centralized model – their Data Integration Access Layer. During an interview in 2022, Mahmoud Yassin, former Lead Architect at ABN AMRO, talked about data mesh. He advocated centralizing offerings in a platform approach that decentralized teams could leverage.
That being said, Mahmoud Yassin believes there’s no one-size-fits-all method or a perfect data architecture. “You should choose what fits your company,” he said. “And also, it’s okay to mix and match.” However, cohesion is crucial.
Journeying to a decentralized model
To become more data-driven and outperform the competition, banks are increasingly considering a decentralized data mesh approach, helping them make better-informed decisions, streamline processes, and enhance the user experience.
Effectively implementing a decentralized model requires a data strategy that blends technology with cultural change. “A lot of data mesh is a mentality shift,” said Mahmoud Yassin. Banks can get started with any data mesh pillar, making it an exciting and tangible prospect, and a way to transform into a truly data-driven enterprise.
Watch the “Data revolution: Superior banking strategies” session here.
For more expert content on industry outlooks and innovation, subscribe to our newsletter or visit our Insights page.