The Acceleration of Data Mesh Adoption in Financial Services: A Strategic Approach

Over the past two years, the financial services sector has seen significant changes in technology and data architecture, especially the insurance, wealth, and asset management subsectors. Organisations are shifting from traditional data warehouses, lake houses, and data factories to the emerging data mesh model. The need for greater scalability, agility, and domain-oriented ownership of data within complex organisations drives this evolution. As financial institutions deal with increasingly complex data landscapes, the data mesh model offers a more decentralised, domain-driven approach that aligns with their business needs while ensuring appropriate standards and more business-oriented guardrails than IT.

The Limitations of Traditional Data Models

Financial services firms have historically relied heavily on centralised data lakes and warehouses. These architectures provided structured environments for storing large volumes of data, but they often became cumbersome, leading to issues such as data silos, unclear ownership, and difficulties in scaling.

Traditional models, such as data factories and lake houses, centralise data into monolithic architectures. While this can be efficient for batch processing and storage, it often hampers agility in real-time analytics, complicates governance, and inhibits teams from making swift, data-driven decisions. Additionally, these models typically depend on a centralised data team, creating bottlenecks that slow decision-making and stifle innovation.

Why Data Mesh? A Strategic Overview

A data mesh architecture decentralises data ownership, positioning data as a product managed by domain experts. This approach is particularly effective in financial services, where business units such as actuarial teams, product management, financial reporting, and operational finance operate across multiple functions.

Instead of centralising data collection, the data mesh model embraces domain-oriented decentralisation, empowering individual teams to manage their own data products. Each domain (e.g., insurance underwriting, asset management portfolios) takes responsibility for the data it generates and consumes. This decentralisation not only makes data more accessible but also increases accountability, ensuring data is high-quality, fit for purpose, and readily available for real-time decision-making.

Strategic Considerations for Operationalising Data Mesh

1. Organisational Readiness and Data Governance: The first step in adopting a data mesh architecture is ensuring the organisation is prepared for such a shift. This involves a cultural change towards decentralised data ownership and accountability within specific domains. Establishing clear governance frameworks that standardise practices while empowering individual teams to operate with agility is crucial. Effective data governance is essential to ensure compliance with industry regulations, such as IFRS 17, for the insurance and wealth management sectors.
2. Interoperability between data domains and topology: Interoperability enables seamless data flow across domains, reducing silos, improving real-time collaboration, and ensuring consistency in shared processes like FP&A and compliance. It also optimises business alignment and data-driven decision-making. Meanwhile, a well-designed domain topology aligns the architecture with the organisation’s structure, ensuring scalability, governance, and operational efficiency. Whether hub-and-spoke or decentralised, the right topology supports agile growth, enhances domain-specific innovation, and maintains necessary governance in regulated environments like insurance and asset management.
3. Data as a Product: Within the data mesh model, data is treated as a product, and each domain assumes responsibility for its data pipelines, ensuring usability, accessibility, and quality. This shift requires organisations to invest in training and upskilling domain teams, enabling them to manage their data environments and adhere to broader corporate governance standards.
4. Self-Service Infrastructure: A self-service data platform is key to successfully implementing a data mesh. For financial services, this means equipping actuarial, risk management, and FP&A teams with tools to curate, process, and analyse data independently. The platform must support on-demand scalability, real-time analytics, and interoperability across systems, allowing domains to operate autonomously without dependence on centralised IT teams.
5. Automated Data Discovery and Lineage: In a heavily regulated sector like financial services, tools that provide automated data discovery, monitoring, and lineage tracking are essential. Ensuring real-time visibility of how data moves across domains is critical for auditability and regulatory compliance, helping avoid the pitfalls of opaque data flows common in traditional centralised models.
6. Data Contracts and Interoperability: In a decentralised data mesh model, domains must communicate via data contracts that define the interfaces and structures for sharing data. Ensuring seamless interoperability across domains (for example, actuarial data integrating with FP&A and risk management data) is vital to the architecture's overall effectiveness.

Designing Infrastructure Based on Topologies

The success of a data mesh implementation depends on selecting the appropriate topology—the arrangement of systems, networks, and processes within the enterprise. In the context of data mesh, two prominent topologies emerge:

1. Domain-Oriented Topology In this model, data is decentralised across business domains. Each function, such as technical accounting, FP&A, and operational finance processes like Procure-to-Pay (P2P) and Accounts-to-Record (A2R), maintains its dedicated data infrastructure. This reduces inter-domain dependencies and allows for rapid, domain-specific innovation.
2. Hub-and-Spoke Topology Here, while data ownership is decentralised, certain centralised services remain to provide governance, compliance, and security. A central governance hub ensures domain teams adhere to common standards, while spokes represent domain-aligned infrastructures that allow flexibility and agility.

Finance Use Case: Optimising FP&A and Operational Finance with Data Mesh

A data mesh architecture can significantly benefit the finance function, including FP&A, business partners, technical accounting, and operational finance processes like P2P and A2R.

Example: Enhancing Financial Planning and Analysis (FP&A)

With data mesh, FP&A teams can leverage real-time, domain-specific insights across the organisation. In insurance, for example, actuarial models, product management, and underwriting data can feed directly into FP&A’s financial models, enabling real-time adjustments to forecasts based on emerging trends.

• Reduced Time to Insight: Traditionally, FP&A teams rely on centralised data, which can be slow to access and process. In a data mesh model, FP&A teams can query underwriting, pricing, or actuarial data directly in real-time, reducing data collection and cleaning time and enabling faster, more accurate decision-making.
• Improved Business Partnering: By directly accessing high-quality, domain-specific data, FP&A teams can engage more effectively with business units such as operations, claims, and investments, improving the accuracy and relevance of insights shared during business partnering discussions.

Example: Optimising Operational Finance (P2P, A2R)

In operational finance processes such as Procure-to-Pay (P2P) and Accounts-to-Record (A2R), data mesh allows finance teams to access data products directly from procurement, treasury, and accounts receivable functions, avoiding reliance on centralised data teams.

• Real-Time Reconciliation: Data from procurement and accounts receivable can be accessed in real-time, enabling quicker reconciliation and speeding up the financial close process.
• Improved Compliance: Decentralised ownership ensures each domain maintains regulatory compliance (e.g., IFRS, Solvency II), reducing the compliance burden on centralised teams and streamlining the audit process.

Benefits for Financial Services: Insurance and Asset Management

1. Enhanced Data Quality: In domains such as technical accounting, underwriting, and claims management, the data mesh model ensures that those with the most expertise in a given area are responsible for the data’s quality.
2. Increased Agility: The domain-driven nature of data mesh enables insurance and asset management firms to adapt more quickly to market shifts, regulatory changes, and operational needs.
3. Stronger Governance and Compliance: The data mesh approach supports robust governance frameworks, ensuring each domain complies with standards without creating bottlenecks within centralised teams.

Conclusion

Adopting data mesh in financial services, particularly within the insurance and wealth management sectors, represents a shift in how organisations manage and operationalise data. Organisations can achieve greater scalability, agility, and alignment with business needs by decentralising data ownership and treating data as a product within specific domains. For finance teams, real-time, high-quality data enables faster insights, enhanced compliance, and improved operational efficiency in key processes such as FP&A, P2P, and A2R. As financial services continue to evolve, the future of data architecture is not centralised—it is distributed, domain-driven, and designed for agility.