Using Disaggregated Regulatory Data to Support Financial Policy and Supervision

Explore CGAP LENS

• Executive Summary
• 1. Introduction
• 2. The Foundations for Segmented Analysis
  • 2.1 Customer Visibility Tiers
  • 2.2 Indicators for Segmented Analysis
• 3. Use Cases for Segmented Analysis across FSA Mandates
  • 3.1 Using Segmented Analysis to Diagnose and Advance Financial Inclusion
• 4. Building Blocks for Disaggregated Data and Segmented Analysis
  • 4.1 Data Strategy
  • 4.2 Data Governance Framework
  • 4.3 High-Quality Disaggregated Data
  • 4.4 Capacity for Segmented Analysis
  • 4.5 Inter-Agency Data Sharing
  • 4.6 Dissemination of Data as a Public Good
  • 4.7 Technology as a Cross-Cutting Enabler
• 5. The Implementation Pathway
  • 5.1 Identifying Your Visibility Tier
  • 5.2 Ensuring Legal and Governance Foundations
  • 5.3 Designing the Data Strategy
  • 5.4 Ensuring proportionality of a phased approach
  • 5.5 Adopting a Collaborative Approach
  • 5.6 Monitoring and Evaluating Progress
• Conclusion
• References
• Annex 1: How to use CGAP LENS
• Annex 2: Questionnaire on Regulatory Data Collection and Use

Executive Summary

Disaggregated regulatory data gives Financial Sector Authorities (FSAs) a clearer view of which customer groups are being served by financial services, under what conditions, and with what outcomes. While aggregate data remains essential for monitoring firm-level and system-level developments, it often hides differences and gaps across customer groups that matter for financial inclusion, consumer protection, and risk management. As regulatory reporting becomes more granular and linkable across datasets, including external ones, FSAs can better diagnose these gaps and vulnerabilities, and decide where targeted and proportionate action is needed. Further, a virtuous cycle can emerge when FSAs share nonconfidential data or analysis with other authorities, Financial Service Providers (FSPs), funders, and civil society. This is not an add-on agenda, but a way to strengthen financial inclusion and financial health policy and while contributing to safe, stable, sustainable, and equitable financial systems.

The conditions for the wider use of disaggregated regulatory data are increasingly in place. FSAs worldwide are moving toward more granular and higher-frequency reporting, supported by SupTech, and broader trends in data and institutional modernization. Recent global initiatives and guidance from AFI, CGAP, GPFI, WFID, IFC-BIS, IMF, We-Fi, World Bank, and the UN underscore, from mutually reinforcing and complementary angles, the value of sociodemographic disaggregation. This creates momentum for FSAs to embed disaggregated data into their policy, regulatory, and supervisory work, where they have clear legal authority to do so, and to adopt a phased approach that ensures proportionality of data collection and protects data privacy and bank secrecy.

This Guide proposes a framework (Figure E1) to help FSAs use disaggregated regulatory data for evidence-based decision-making by strengthening customer visibility, enabling segmented analysis, and building the institutional foundations to support its effective use. The focus is on data reported by regulated FSPs ("regulatory data"). Demand-side surveys and other supply-side data remain essential complements to regulatory data but fall outside of this Guide's scope.

The Customer Visibility Tiers serve as a guide for FSAs to understand how extensively their institution is currently using sociodemographically disaggregated regulatory data, and where they have the potential to do more. As visibility increases, so does the scope for segmented analysis across mandates, to identify differential trends and gaps. This feeds an evidence-based decision cycle in which FSAs can design more targeted policies, and regulatory or supervisory responses, while monitoring distributional progress.

Source: Authors

Every FSA, at every stage of data maturity, has a viable next step toward higher customer visibility. Strong institutional building blocks help ensure that disaggregated data is collected and used in ways that are safe, reliable, and actionable, enabled by technology. A phased and proportional implementation pathway helps FSAs assess their current capacity, prioritize reforms, and sequence next steps. This Guide, together with CGAP LENS—a repository of more than 200 indicators that can be broken down across several dimensions including sociodemographic—is designed to help FSAs identify that next step.

This Guide is primarily written for FSAs—especially their financial inclusion departments, as well as their data, IT, supervisory and research teams—and may also help other stakeholders. While financial inclusion teams within FSAs may already appreciate the value of sociodemographic disaggregation, supervisory and statistics teams may lack full awareness of its relevance to their work. A shared institutional understanding can help FSAs reduce informational silos, reuse data across departments, and reduce duplication of reporting requests to FSPs.

The Guide helps enhance institutional awareness of the value of disaggregated data, illustrating use cases across key mandates, including financial inclusion; consumer protection; stability, safety and soundness; sustainability; and market development. For financial inclusion, the use cases described in this Guide are well established and action oriented, whereas for other mandates, they are largely exploratory and intended to stimulate thinking and research rather than prescribe action. Each FSA must assess the evidence generated by disaggregated data and decide whether and how to act, based on its legal mandates, strategic priorities, and other contextual factors. Other authorities may use the Guide to align data reforms with broader financial sector objectives, whilst development partners and funders may use it to design technical assistance and investment strategies. FSPs may use it to understand emerging reporting trends, and researchers, advocates, and other public-interest actors may use it to interpret and use disaggregated evidence responsibly.

Seven institutional building blocks support FSAs' collection and use of sociodemographically disaggregated data (Figure E1). Together, these building blocks help ensure that disaggregated data are not only collected but also managed, analyzed, and shared, in ways that are safe, reliable, proportionate, and actionable, while protecting data privacy. A clear data strategy is especially important to define priorities, sequence reforms, manage costs, and ensure that disaggregation is embedded in an FSA's broader data modernization efforts rather than added as a parallel reporting layer. Robust data governance is equally essential to clarify responsibilities, establish common standards, support data quality and interoperability, and set the rules for access, sharing, retention, and responsible use of granular data. Privacy protection and proportionality are core design and implementation principles, ensuring that any expansion of data collection or use is grounded in clear legal authority, robust safeguards, and a strong case for serving the public interest.

The case for FSA action is clear, and implementation should be phased and proportionate. FSAs do not need to move immediately toward the most advanced data model. They can start with a realistic assessment of their current Customer Visibility Tier, using the questionnaire provided in this Guide, then use CGAP LENS and the use cases presented in here to prioritize relevant indicators and analysis and build the necessary building blocks over time. Incremental rollout—starting, for example, with credit, higher-capacity FSPs, and/or a selected set of indicators disaggregated by one or two sociodemographic dimensions (such as gender and age) before expanding—will often be more realistic and safer than overly ambitious roadmaps. As FSAs improve their capacity to collect and analyze disaggregated records, they can disseminate nonconfidential data and analytical findings as a public good, enabling policymakers, development partners, funders, FSPs, researchers, advocates, and other actors to identify gaps, design or fund reforms, benchmark progress, improve products, strengthen accountability, and contribute to more inclusive, responsible, and resilient financial systems.

1. Introduction

Sociodemographic disaggregation in the financial sector data is attracting growing interest. This is mostly driven by a recognition that whilst aggregate indicators are essential, they are insufficient for understanding how different groups of individuals and firms experience financial services, which is a necessary input for designing effective and equitable policy responses (Alonso and Dezso 2024). A number of recent developments show the attention that this is getting from the international community. The Global Partnership for Financial Inclusion (GPFI) and the World Bank have recently called on financial sector authorities (FSAs) to collect and use data disaggregated by sociodemographics as a foundation for evidence-based action (GPFI 2025). The International Monetary Fund (IMF) is revamping its Financial Access Survey (FAS) to expand the production of granular indicators, disaggregated by gender, income, age, and living area across more than 190 reporting economies, including for micro, small and medium enterprises (MSMEs). Adoption of the Women Entrepreneurs (WE) Finance Code, which promotes gender-disaggregated data collection and usage, continues to gain global momentum (We-Fi 2025a), and recent guides from the United Nations (UN), the Alliance for Financial Inclusion (AFI) and the Women's Financial Inclusion Data (WFID) Partnership provide practical tools for integrating gender and other sociodemographic dimensions into financial sector data systems (UN Women 2026; AFI 2024, FAW & Data2X 2023). Together, these developments point to a growing international consensus that better supply-side disaggregated data on individuals and MSMEs is not a technical refinement but a foundational requirement for more effective, targeted, and inclusive actions and a concrete contribution to the commitments embedded in Sustainable Development Goal 5 on gender equality and women's economic empowerment (UN 2015).

This Guide focuses on regulatory data, which constitutes the lion's share of supply-side financial data and the lever most directly under FSAs' control. Regulatory data are one of several complementary supply-side sources in the financial sector data landscape, alongside data held by financial services providers (FSPs), market infrastructure data, social media monitoring data, market intelligence surveys, and demand-side surveys such as the World Banks' Global Financial Inclusion Database (Findex), FinMark Trust FinScope surveys deployed across several countries, as well as dedicated national instruments. While demand-side data have unique advantages—they capture financially excluded populations, informal behaviors, and self-reported outcomes—regulatory data offer wide coverage of regulated activities in a standardized manner and at higher frequency, greater granularity, and lower cost, allowing FSAs to observe behaviors and outcomes as they unfold.

High-quality regulatory data are crucial for effective financial sector policy, regulation, and supervision, including in support of inclusive finance (IFC-BIS 2018). FSAs, including in emerging markets and developing economies (EMDEs), use them to monitor market dynamics and risks, measure outcomes, and take evidence-based action to promote stability and soundness, responsible financial inclusion, and sustainable market development. Through segmented analysis using disaggregated regulatory data, FSAs can identify systematic differences in outcomes across population groups, detect vulnerable segments, and act with greater precision across their mandates (GPFI 2025; IFC-BIS 2024; IMF 2022a).

Many FSAs are already modernizing their regulatory reporting frameworks, moving toward more granular, higher-frequency data, in a shift that is being accelerated by supervisory technology (SupTech) and artificial intelligence (AI) (BoE 2026; ECB 2017; IFC-BIS 2017, 2021, 2024; Moreira et al. 2026; Restoy 2025). Authorities including the Central Bank of Brazil, the Financial Superintendence of Colombia, the National Bank of Rwanda, the Bank of Thailand, and the Monetary Authority of Singapore already have granular datasets for at least some types of regulated activities. This broader modernization creates a key opportunity to embed sociodemographic disaggregation—by gender, age, location, income, and other attributes—into regulatory reporting and analysis as part of key data reforms.

Gender disaggregation is the most consistently documented and analytically tractable starting point for sociodemographic disaggregation. Across countries—particularly EMDEs—gender-disaggregated data reveal that women have worse experiences and lower benefits from financial services, even where formal access gaps have narrowed (AFI 2024; IMF 2024; UN Women 2026). Where gender identifiers are available, FSAs can produce and analyze gender-disaggregated indicators, benchmark against other countries (e.g., using IMF FAS data), and inform targeted action. The capacity for segmented analysis can be extended from a focus on gender to encompass other sociodemographic dimensions of interest (e.g., age, location), enabling progressively more intersectional analysis over time. For these reasons, gender serves throughout this Guide as the primary illustrative entry point.

FSAs' interest in diagnosing and monitoring diversity and inclusion (D&I) is also growing, reinforcing the need for sociodemographic data from both customers and employees of FSPs. FSAs are increasingly taking an integrated view of D&I that covers FSP governance and staffing as well as how FSPs serve and include their customers (BIS 2022b). FSPs' exclusionary practices can create consumer, reputational, and legal risks, while workforce diversity can support more balanced decision-making by reducing groupthink and may strengthen economic competitiveness and financial value by enhancing talent attraction, innovation, customer understanding, and long-term resilience (Edmans et al. 2024). However, a full diagnostic requires sociodemographically disaggregated data that most FSAs do not yet systematically collect. Mandatory disclosure and regulatory reporting requirements on D&I metrics are key levers to close this knowledge gap.

The central objective of this Guide is to help FSAs integrate gender and other sociodemographic dimensions into regulatory data as a standard practice, strengthening diagnostics and informing evidence-based action. The Guide provides a practical framework for leveraging disaggregated data, drawing on the literature, CGAP's own work, consultations with FSAs and international experts, and direct technical collaboration with select FSAs between 2024 and 2025. It uses terminology consistent with the Irving Fisher Committee on Central Bank Statistics (IFC-BIS 2017, 2024), extended to sociodemographic breakdowns in regulatory data (see Glossary). The Guide is primarily written for FSAs and is intended to support cross-functional engagement across financial inclusion, data and statistics, IT, policy, supervision, and research teams. It is designed both for staff who need to understand and make the institutional case for disaggregated regulatory data and for those who are already convinced of its value but need practical tools to institutionalize its collection, analysis, and dissemination.

The Guide may also support policymakers and other public authorities in aligning data reforms with broader financial sector objectives; development partners and funders in designing technical assistance and investment strategies; FSPs in understanding emerging reporting trends; and researchers, advocates, and other public-interest actors in using disaggregated data responsibly.

This Guide is accompanied by CGAP LENS, an interactive catalog of over 200 indicators spanning five regulatory mandates across six financial service areas. CGAP LENS helps FSAs identify relevant indicators, breakdowns, and exploratory questions for sociodemographic analysis.¹

The Guide is organized in six main sections. Following this Introduction, Section 2 presents the four Customer Visibility Tiers of regulatory data and CGAP LENS. Section 3 illustrates how FSAs can conduct segmented analysis to inform evidence-based action across institutional mandates—financial inclusion; consumer protection; stability, safety, and soundness; sustainability (including diversity); and market development. Section 4 describes seven building blocks for integrating sociodemographic dimensions into regulatory data as a standard practice, with particular attention to the challenges and opportunities of granular data. Section 5 provides key considerations for implementation, including a self-assessment tool for FSAs to identify their current Customer Visibility Tier and guidance for a proportional and collaborative approach. Section 6 concludes with a call to action for FSAs.

2. The Foundations for Segmented Analysis

2.1 Customer Visibility Tiers

The scope of segmented analysis an FSA can undertake depends on the granularity of its regulatory data and its ability to link customer-level records with other databases, within the FSA and externally. Beyond typical breakdowns by product, channel, or provider type, sociodemographic analysis requires granularity at the individual customer level. There are four customer visibility tiers reflecting the maturity of an FSA's data infrastructure and the granular data analysis that it enables (Figure 1). The tiers are cumulative, with each higher tier embedding the capabilities of those below it and expanding the analytical scope and depth through an increasingly integrated and granular data framework.

Source: Authors

The tiers are as follows:

Tier 0: Sociodemographic Blindness. Data are collected at the portfolio level, segmented by product, channel, or customer type (i.e., individual, MSME). Some data may be collected in granular form at the account or transactional level, and possibly connected to external databases (triangulation), but without sociodemographic information (e.g., gender, age) or a customer-level identifier that would allow the FSP to link records and obtain sociodemographic traits. Customer-segmented analysis is therefore not enabled.

Tier 1: Aggregate Sociodemographic Breakdowns. Regulatory data include at least one sociodemographic trait (e.g., gender), on top of Tier 0 breakdowns. Customer-segmented analysis is enabled, including through some triangulation with external data, yet limited to these predefined categories. FSAs can identify high-level gaps (for example, women's underrepresentation among retail borrowers) that may trigger policy actions.

Tier 2: Intersectional Aggregates. FSPs report granular data with customer-level sociodemographic variables (e.g., gender, age, income, residence, marital status, ID number) for at least one dataset—for instance, account ownership data or volumes and values of loans. Tier 2 enables FSAs to develop custom segments directly from the granular data collected, including intersections across two or more sociodemographic traits. For example, FSAs may detect whether young rural women face lower access to credit or savings, higher rejection rates in loan applications, or higher interest rates on loans. FSAs can also understand how gaps vary across intersectional subgroups (e.g., distribution of gender gaps across ages and locations). Tier 2 does not require the existence of a unique ID number per customer. Yet when this is available, it can help increase data quality (e.g., eliminate double counting) and it may allow linkages to external databases, which may contribute to further enhancing the scope and breadth of the analysis.

Tier 3: Integrated Microdata (Customer 360). FSPs report granular data across most datasets, including transactional data for the financial products under the FSA's purview, and every dataset includes a unique customer identifier, such as a national ID number, for each customer record. This allows the FSA to link a single individual's records across multiple products and providers, creating a "customer 360" view, and tracking whether genuine progress is being made in financial inclusion; for example, whether women who gained access to credit also show improvements in savings or have experienced increased fraud or difficulty repaying credit and, eventually, whether these findings change by age group or geography.² Tier 3 effectively enables a shift from a focus on average gaps in certain subgroups to understanding the depth and dynamics of financial inclusion and behavior across multiple dimensions (Box 1).

Analytical capabilities at Tier 2 and Tier 3 expand significantly when FSAs can connect individual customer records to external databases—e.g., credit bureaus, tax registries, collateral registries, or social protection systems. Such linkages enable FSAs to observe factors outside the regulated financial sector, such as existing liabilities, collateral availability, or income sources, deepening diagnostic analysis and impact assessment. The FSA's responsibility for data protection, security, and confidentiality increases as data become more granular and cover more sensitive information (e.g., savings, insurance). Linkages to external databases require clear legal justification for processing sensitive personal data, as well as stronger governance and privacy controls (see Section 4).

2.2 Indicators for Segmented Analysis

Segmented analysis must be grounded in indicators that can be broken down by customer group, FSP, product, and channel types. For example, the share of outstanding retail debt in delinquency could be calculated overall, by loan type, and—where borrower sociodemographic data are available—disaggregated by gender, age and location to reveal differences such as those between young urban and young rural women. The Center for Microeconomic Data at the Federal Reserve Bank of New York and the Central Bank of Colombia are among the FSAs already producing this kind of intersectional analysis. The former tracks household debt and credit indicators by borrower age and state based on their Consumer Credit Panel built using credit reporting data, while the latter examines differences in credit quality, nonperforming and risky loans, and loan interest rates between men and women (Federal Reserve Bank of New York 2026; Banco de la República 2023, 2026).

CGAP LENS³ helps FSAs identify indicators that are suitable for segmented analysis. It is a free, interactive tool designed for FSAs and other financial market actors seeking to explore and strengthen their use of disaggregated or granular regulatory data (CGAP 2026). It complements this Guide by providing references for indicators for disaggregated reporting and analysis, helping FSAs decide which data or disaggregated indicators to request from FSPs and which ones to share with the wider market, helping shape the data architecture. The tool provides a curated catalog of over 230 indicators spanning five policy mandates (financial inclusion; consumer protection; stability, safety, and soundness; sustainability; and market development) across six financial service areas: payments, credit, insurance, investments, pensions, and savings (Figure 2). CGAP LENS is most useful when it is applied to specific policy questions. It helps users explore the intersection of policy mandates and financial services, for example, identifying which consumer protection indicators apply to segmented analysis for credit markets, or which financial inclusion indicators are most relevant to payments, enabling authorities to target their data collection efforts with greater precision.

A key feature of CGAP LENS is the ability to build and export custom indicator sets tailored to a user's specific priorities and data capacity. Each indicator includes definitions, data requirements and limitations, suggested breakdowns to support segmented analysis, as well as exploratory questions considering sociodemographic nuances. When relevant, the tool also maps relationships to indicators proposed by other global initiatives, including AFI, G20-GPFI, IMF FAS, and WE Finance Code. Users can create multiple named collections, annotate them, and export their selections as structured reports, supporting the practical work of designing or refining data collection frameworks.

Source: Authors

3. Use Cases for Segmented Analysis across FSA Mandates

3.1 Using Segmented Analysis to Diagnose and Advance Financial Inclusion

Financial inclusion is a central mandate for FSAs in many countries (BIS 2021). Yet, aggregate indicators can mask deep structural disparities that can hinder progress in achieving it. Box 2 illustrates this idea for Rwanda, and the rest of this section explains the types of segmented analysis that FSAs can conduct to assess and strengthen their financial inclusion analysis, exploring how findings can be translated into concrete policy action.

3.1.1 From Data to Evidence: Generating Actionable Insights Using Segmented Analysis

For FSAs with a financial inclusion mandate, segmented analysis should be a continuous resource (not a one-off exercise) supporting mutually reinforcing types of analysis that pay special attention to groups facing heightened barriers (for example, low-income groups, older generations, or women). In practice, segmented analysis supports three linked analytical functions:

• Establish a baseline. The first step is a point-in-time assessment of gaps and distributional differences across priority customer groups, using indicators disaggregated by key sociodemographics of interest, customer type, provider type, and product type. This baseline identifies where disparities are most pronounced, guides prioritization of follow-up analysis, and creates a reference point against which future interventions can be assessed.
• Monitor changes over time. Once a baseline is in place, segmented analysis enables ongoing tracking of how disparities evolve, early detection of emerging gaps, and identification of areas requiring further investigation or action.
• Assess outcomes and impact. The third function is to evaluate whether specific interventions have produced their intended outcomes (Lahaye et al. 2024; Nielsen and Meckler 2026). Comparing post-intervention segmented data with the baseline and subsequent trends allows FSAs to draw evidence-based conclusions about what works, for which groups, and under what conditions.

Together, these three functions help FSAs generate actionable evidence across interconnected financial inclusion dimensions (access, uptake, usage, quality and outcomes)⁴, as follows:

1. To assess access inequalities, FSAs can use data on the distribution of physical access points (e.g., branches, agents, ATMs, merchant points of sale) disaggregated by geography and, where available, by the gender or ownership profile of access point operators. This allows FSAs to identify underserved areas and assess whether agent and merchant networks reflect the diversity of the population they serve. For example, in certain contexts, women agents or FSP staff may help overcome structural barriers that limit women's comfort and ability to access services.
2. To diagnose uptake gaps and barriers, FSAs can look at account ownership and first-time product uptake rates, disaggregated by gender, age, and geography, revealing whether certain groups are adopting formal financial services and products at lower rates and whether gaps are widening or narrowing. Tracking trends in new account openings, or with first-time borrowers or policyholders, by customer segment and product type helps FSAs assess whether financial inclusion strategies are translating into measurable progress for priority groups.
3. To understand usage and dormancy patterns, FSAs can analyze active account rates, transaction frequency, and average credit or deposit balances by customer type segment to distinguish nominal inclusion from meaningful usage. High dormancy or declining usage among specific groups can signal that products are not meeting customer needs, due to barriers such as fees, limited digital literacy and financial capability, or product design.
4. To identify product quality and conduct risks, FSAs can analyze disaggregated indicators such as average cost-to-customer, rejection rates, claims ratios, complaint incidence, and repayment performance, and assess whether some customer segments systematically face worse conditions. In credit markets, differences in annual percentage rates (APRs), collateral requirements, or product approval rates across gender and/or age groups may reflect legitimate risk differentials, but may also point to discriminatory underwriting, overly rigid eligibility criteria, or product design that disadvantages particular segments (BIS 2022b). Where groups with lower approval rates also show stronger repayment performance, this may indicate unjustified differences in product terms that warrant further supervisory examination.
5. To assess whether financial inclusion is supporting positive customer outcomes, FSAs can use indicators linked to financial health or well-being, including the ability to manage financial needs and obligations (e.g., incidence of overdrafts, fraud losses, sustainable debt service ratios), cope with risk (e.g., active insurance coverage, liquid savings buffers), and pursue goals and opportunities (e.g., participation in long-term savings or investment) (Nielsen et al. 2026). Quality and outcomes indicators connect directly to the consumer protection mandate. Where financial inclusion analysis points to potential conduct risks, it can be complemented by the supervisory analysis described in Consumer Protection (Section 3.2). Quality indicators serve as proxies for whether consumer protection principles (e.g., transparency, fairness, suitability, recourse) are followed in practice.

Consumer protection frameworks are evolving from a compliance focus toward an outcomes orientation, whereby FSAs look beyond FSP policies and procedures to their actual effects on consumers, making disaggregated quality indicators a natural input for assessing consumer protection outcomes. Outcomes indicators associated with financial health or well-being operate at a higher level, reflecting people's money management ability, resilience, planning capacity, and confidence, which are shaped by financial and non-financial factors; some outcomes may only be measured through proxy indicators of consumer risks with significant welfare implications.

Table 1 maps these five financial inclusion dimensions to key policy and exploratory questions, and to examples of CGAP LENS indicators that can be disaggregated by gender, age, location, product type, and other traits to support segmented analysis. CGAP LENS provides a fuller catalog of indicators, suggested breakdowns, and exploratory questions that FSAs can draw on depending on their mandate, policy goals, and analytical priorities (see Section 2 and Annex 1).

Table 1. Identifying the Right Levers to Inform Customer-Segmented Financial Inclusion Analysis

Source: Authors

3.1.2 From Evidence to Action: How FSAs Can Advance Financial Inclusion Using Segmented Analysis

The actions available to an FSA in response to insights from customer segmented analysis depend on its mandates and objectives, its strategic priorities, and the tools at its disposal. Experience across countries shows that disaggregated data can strengthen regulatory and supervisory actions, inform national strategies and policy targets, and influence how FSPs design products, assess risk, and target outreach (Box 3).

Taking regulatory and supervisory action

A key objective of segmented analysis is to inform regulatory and supervisory action, including in connection with consumer protection and prudential oversight (AFI 2024). Where gap analysis reveals structural barriers embedded in regulation (e.g., collateral, documentation, or Know your Customer (KYC) requirements that disproportionately disadvantage specific groups), regulatory adjustments may be needed. Segmented evidence can also support supervisory action where it reveals persistent disparities in pricing, rejection rates, complaints, or repayment performance across customer segments. There are several examples of FSA action on this front:

• In Mexico, the National Banking and Securities Commission (CNBV) reduced loan loss provisioning requirements for certain loans to women after gender-disaggregated data analysis showed higher spreads and lower average disbursements despite better repayment rates. As a result, the share of personal loans directed to women increased, interest rates decreased, and average credit amounts rose (Becerra-Ornelas et al. 2024).
• In Pakistan, disaggregated data revealed a large gender gap in account ownership accompanied by very low female participation among bank staff and branchless banking agents, prompting the State Bank, under its Banking on Equality Policy, to require branchless banking providers to increase the share of women agents and embed a gender lens in product design, marketing, and staffing (SBP 2021a, 2021b). These actions contributed to women's account ownership rising from 39 to 52 percent between 2021 and 2025, with the gender gap narrowing by 9 percentage points (Ministry of Finance 2026).
• In Colombia, years of Financial Inclusion Reports documenting persistent gender gaps motivated the SFC to introduce a new regulatory report in 2021 requiring credit providers to submit active interest rates disaggregated by gender and other customer traits. The SFC is now further upgrading its credit registry infrastructure through the MURIC project. These efforts are deepening the collection and analysis of granular, sociodemographically disaggregated credit data (SFC and Banca de las Oportunidades 2025; Banco de la República 2023).

More broadly, segmented analysis supports risk-based prudential and conduct supervision by identifying the FSPs and market segments where disparities or segment-specific risks are most pronounced, including pricing disparities, differential rejection rates, discriminatory underwriting practices, concentrated arrears, or high dormancy among particular groups (see Consumer Protection in Section 3.2).

Taking and informing policy action

Sharing non-confidential disaggregated data and findings with government counterparts, development partners, investors, FSPs, and the public, allows FSAs to shape policy design, benchmark progress, and strengthen accountability. One important channel is the design and monitoring of national financial inclusion strategies (NFISs). Segmented analysis helps authorities identify which groups are being left behind, set more calibrated targets, and ensure those targets are measurable over time (World Bank 2021). There are several examples of FSA action on this front:

• In Tajikistan, supply-side data showing that only 21 percent of loan amounts disbursed in 2020 went to women informed the 2022 NFIS, which included a dedicated pillar on women and youth empowerment and assigned responsibility for gender-disaggregated data collection and monitoring to the NFIS Executive Committee (National Bank of Tajikistan 2022).
• In Bangladesh, the central bank required 10 percent of total MSME loan portfolios to be allocated to women entrepreneurs in 2019, raising the target to 15 percent in 2024 (Bangladesh Bank 2019).
• In Pakistan, successive NFISs have progressively tightened women's financial inclusion targets, most recently aiming to reduce the gender gap in account ownership from 34 to 25 percent by 2028 (SBP n.d.).
• In Nigeria, the NFIS prioritizes women's financial inclusion through dedicated working groups, lending goals for women and women-owned enterprises, and gender-disaggregated monitoring indicators (World Bank 2018).

Disaggregated analysis can reveal structural barriers for certain groups and inform targeted support programs, such as credit guarantee schemes or refinancing facilities.

3.2 Using Segmented Analysis to Support Other Mandates

Regulatory disaggregated data collected to support analysis under one mandate do not need to be collected again for others. Once in place, they should become a shared asset that teams across the FSA can draw on to deepen their own analysis. For example, financial inclusion teams would use data on NPL from prudential returns and, conversely, stability reports could rely on credit uptake data collected by the financial inclusion department. Realizing this potential requires key building blocks such as an adequate data governance framework including data access and sharing protocols, appropriate privacy safeguards, and technology infrastructure that makes the data accessible to different teams without duplicating reporting requirements (see Section 4.2). This "collect once, use widely" logic is one of the strongest arguments for investing in granular data infrastructure.

The justification for segmented analysis, however, differs by mandate. For financial inclusion, sociodemographically disaggregated data are foundational to understanding and assessing who is being served and who isn't. For other mandates (consumer protection, financial stability, sustainability, and market development), segmented analysis plays a complementary role: it enriches existing analytical frameworks by revealing whether risks, vulnerabilities, or market dynamics differ across population segments. An FSA's stability team, for example, may find that granular credit data collected for financial inclusion monitoring also reveal the concentration of over-indebtedness among specific segments, relevant to prudential risk assessments. Likewise, a consumer protection team may use the same data to identify customer segments paying higher loan processing and application fees and having high NPLs, raising unfair treatment concerns. In both cases, the segmented lens deepens and adds a distributional dimension to analysis that would otherwise remain at the aggregate level.

Analyzing progress in financial inclusion with a sociodemographic lens can thus be seen as a practical starting point: it helps generate disaggregated data that other teams can leverage for their own mandates and consolidate institutional interest in customer-centric, segmented diagnostics. As FSAs progress to higher Customer Visibility Tiers, the range and depth of cross-mandate applications expand accordingly. Table 2 maps CGAP LENS indicators with exploratory entry points for gender and other sociodemographic analyses across consumer protection, financial stability, sustainability, and market development mandates.

Table 2. Identifying the Right Questions to Inform Customer-Segmented Analysis in Support of Mandates other than Financial Inclusion

Source: Authors

Consumer protection

Segmented analysis can help FSAs improve the identification, assessment, and mitigation of consumer risks and negative consumer outcomes in the financial sector, by analyzing patterns across gender and other demographic dimensions. Potential uses of this include: identifying gender and intersectional gaps in the incidence of consumer risks and outcomes (e.g., over-indebtedness, fraud losses) at market and FSP levels; refining the definition and monitoring of customer vulnerability; and understanding the linkages between FSP gender diversity and business conduct and overall consumer risk profiles.

Segmentation by sociodemographic groups is essential for supervisors willing to add a customer-centric approach to their consumer protection frameworks (Izaguirre 2020). Segmentation helps FSAs strengthen risk-based conduct supervision by identifying which consumer groups are being underserved, overcharged, or treated unfairly, by which FSPs and in what product and channel segments—both through the supervision of individual institutions and through market monitoring (Izaguirre et al. 2022). Negative outcomes may disproportionately harm women, minority genders, low-income consumers, and other segments in vulnerable situations (Duflos et al. 2024).

FSAs can conduct segmented analysis of data on FSPs' direct engagement with customers (customer interactions and operations). Complaints data are a particularly accessible and high-value starting point for FSAs to identify potential differences in complaint types, or in resolution outcomes, across customer groups. Patterns in complaint types, especially those related to fraud risks or misconduct such as undisclosed fees and misselling, can signal risks concentrated in specific customer segments by FSP and product type. For insurance, a sociodemographic analysis of insurance claims ratios and processing times by policy type can reveal whether certain consumer segments are systematically sold low-value policies with poor payout rates. Analyzing application rejection rates by customer gender, income, age, location, and other relevant sociodemographic criteria can also help FSAs determine whether specific customer segments face disproportionate barriers to taking up products, and whether this reflects genuine risk differentials or potential bias.

FSAs can also conduct segmented analysis of financial data that reveals consumer risks and outcomes from using financial services. Analyzing processing and administration fees by customer segment helps FSAs identify potential unfairness in product design, pricing conditions or transparency, for example, whether certain groups are systematically charged higher fees for equivalent products or services. In pension markets, analyzing fund transfer patterns, particularly transfers from higher to lower-returning funds, by customer gender, agent channel, and geography can reveal whether vulnerable customers are being steered toward less advantageous products by agents responding to fee structures that incentivize misconduct. In credit markets, data can surface discriminatory patterns; for instance, women receiving substantially lower credit than men, despite being just as good at repaying.

If the findings from the customer segmented analysis point to potentially significant vulnerabilities or risks, FSAs could consider further enquiry and/or action. Examples for action include:

• Investigate and address identified conduct risks through supervisory engagement, including thematic reviews, inspections, and corrective measures such as agreed-upon action plans for FSPs to improve practices, change product terms and conditions, and withdraw or modify marketing materials. For example, the Central Bank of Ireland conducted a thematic review of diversity and inclusion across insurance firms, including analysis of gender gaps in remuneration, policies, procedures and practices (Central Bank of Ireland 2020).
• Refine the use of "customer vulnerability" as a driver of risk-based supervisory prioritization, by combining key sociodemographic traits such as gender, age, income, or geographic data to direct thematic reviews and inspections toward the segments most exposed to consumer risks. In Brazil, for instance, there is a draft law requiring FSPs to offer additional transaction confirmation steps for elderly customers who sign up for this enhanced protection measure, since they are a common target of scammers (ICL 2026).
• Publish disaggregated market data on pricing, complaints outcomes, and loan and account portability patterns to increase transparency, reduce information asymmetries, and empower vulnerable consumers to make better-informed choices. For example, the U.S.'s Consumer Financial Protection Bureau (CFPB) published annual consumer response reports that include complaint statistics by special population segments (elderly and servicemembers) and geographic region (CFPB 2025).
• Engage FSPs proactively with analytical findings, signaling regulatory expectations and raising awareness of the legal, conduct, and reputational risks associated with practices biased by gender and other sociodemographic characteristics. South Africa's Financial Sector Conduct Authority (FSCA) engaged closely and collaboratively with FSPs to improve their regulatory reporting requirements of conduct indicators and include segmented data to support an outcomes-oriented supervisory approach (Koning et al. 2022).
• Use enforcement powers to penalize and deter further misconduct. Where misconduct is confirmed for a group of customers, FSAs may use enforcement powers such as issuing warning letters and imposing sanctions such as fines and limitations to FSP operations. For example, the Australian Securities and Investments Commission secured major financial penalties against FSPs for failing to respond to borrowers who were facing financial hardship, as well as remediation payments to affected customers (ASIC 2026).
• Adjust product and conduct regulations. Where sector-wide harmful practices are identified, policy and regulatory reforms may be warranted to protect vulnerable segments, including imposing stricter suitability and affordability requirements (e.g., pricing), prohibiting product features that systematically produce poor outcomes, or mandating enhanced disclosures at the point of product switching or data consent. Mexico's pension regulator (Consar), for example, strengthened regulation to address agent misconduct toward vulnerable segments and improve the security of customer interactions and transactions (CGAP 2022b).

Stability, safety, and soundness

Most prudential data are classified at Visibility Tier 0 as they are designed to monitor FSP- and system-level risks. As a result, sociodemographic analysis is generally not possible within prudential supervision in most jurisdictions. Disaggregated data could, however, enable another largely unexplored line of inquiry: whether the gender diversity of FSP leadership and staff has implications for prudential risks, from credit performance to deposit stability. A growing body of literature suggests that greater gender diversity within FSPs may be associated with stronger corporate governance, more effective risk management, improved financial performance, and better management of climate-related financial risks (Andries et al. 2017; Palvia et al. 2015; Menicucci and Paolucci 2022; Sahay et al. 2017; Sahay and Čihák 2018; FRB 2015; BIS 2022a).

For credit risk, gender-disaggregated Nonperforming Loans (NPLs), delinquency, and loan restructuring data can help FSAs assess whether certain customer segments are systematically associated with higher or lower credit risk. This is increasingly important as digital credit expands globally and as credit outcomes become central to financial health analysis. This analysis can begin with a small set of core questions: Are there gender disparities in delinquency and NPLs across different loan types and sizes? Are FSPs with a higher share of women in their customer base more or less likely to exhibit healthier loan portfolios? How does loan performance after restructuring vary across customer segments?

For operational risk, segmented analysis can help FSAs identify legal and reputational risk channels that may have prudential relevance. In addition to the fraud patterns discussed above, practices that disproportionately expose certain customer groups to fraud, misconduct, or exclusion may generate complaints, litigation, and loss of trust, all of which can impact operational and conduct risks.

For insurance, the excessively low claims ratio mentioned earlier, when concentrated among vulnerable customer segments, could signal long-term strategic risk for insurers, as this can lead to reputation risk or even legal risk as consumer advocacy groups may challenge FSP conduct in court.

At the system level, disaggregating deposit flow and volatility data by customer demographics such as gender or age, account type, and FSP characteristics can reveal whether female and/or aged customers exhibit more stable saving behaviors and whether FSPs with greater gender diversity in their customer base or workforce tend to display higher deposit stability.

If findings are deemed significant, FSAs may consider a range of responses which could include:

• Examine the relevance of sociodemographic variables for risk-based supervision frameworks, exploring whether gender, age, geography, and other relevant variables may serve as potential risk factors or mitigants when assessing FSP risk profiles.
• Explore whether findings on segment-level risk profiles could inform proportionate adjustments to supervisory or regulatory frameworks. Where analysis consistently shows that certain customer segments (e.g., women-led MSMEs, women borrowers) are associated with lower risk and stronger financial performance, FSAs may consider adjusting regulations or supervisory expectations.
• Use segmented findings to strengthen supervisory dialogue on legal and reputational risk, including where conduct patterns affecting specific customer groups may expose FSPs to legal claims, reputational damage, loss of trust, and broader operational or franchise risk.
• Enhance macroprudential monitoring by using disaggregated data to reveal heterogeneous risk exposures, vulnerabilities, and financial behaviors across different population groups, sectors, regions, or firm types.
• Publish disaggregated findings to increase market transparency, reduce information asymmetries, and signal to FSPs the FSA's supervisory expectations for managing sociodemographic risks.

Sustainability (ESG)

Disaggregated data can reveal whether the expansion of green and sustainable financial products is reaching vulnerable segments, such as low-income women, female smallholder farmers, and women-led MSMEs, or remaining concentrated among large corporate clients. Combining sociodemographic data with geographic and portfolio information allows FSAs to assess whether climate-related regulations are inadvertently excluding vulnerable groups, and whether FSPs with greater gender diversity are more effective at serving these segments. Actions can include coordinating with the public and private sectors to close climate insurance coverage gaps for vulnerable populations, adjusting capital markets regulations to ease MSME access to green bond issuance, and publishing disaggregated findings to set clearer FSP expectations around inclusive sustainable finance.

Sociodemographically disaggregated data can also help FSAs assess diversity gaps and inform action both in the financial sector and in their own institutions. Diversity can be relevant to FSAs' core objectives by strengthening governance and helping to reduce reputational, conduct, and legal risks in regulated financial institutions (BIS 2022b; IAIS 2022). FSAs can monitor trends in gender representation and pay gaps across FSPs, identify which types of FSPs or functions present the greatest disparities, and benchmark results against peer jurisdictions. Chile illustrates this well: CMF data showing that women held less than 20 percent of managerial positions and less than 10 percent of board seats across most FSPs informed both the financial sector gender KPIs of a USD 3 billion sustainability-linked sovereign bond and a 2025 law capping single-sex board representation at 60 percent by 2032 (CMF 2022; Alonso 2025; Ministerio de Hacienda 2023; Ministerio de Economía, Fomento y Turismo 2025).

Actions FSAs can take include:

• Share or publish aggregated data or findings on gender diversity and pay gaps at FSPs, making data visible to gender advocates, policymakers, and the public—as the European Banking Authority does through its annual diversity benchmarking reports and NBR has recently started to do with the Women in Financial Services report (EBA 2026; NBR et al. 2026).
• Convene the industry to discuss findings, identify root causes, and co-develop short- and long-term solutions for closing gender gaps.
• Review internal human resources policies to address structural barriers, including adopting inclusive recruitment practices, mentorship programs, flexible work arrangements, transparent pay policies, and anti-discrimination training.
• Integrate gender diversity into supervisory frameworks, treating it as a potential risk factor in risk assessment frameworks.
• Act on identified internal biases: Where data reveal gender bias in licensing or supervisory decisions, FSAs could commission an independent bias assessment, deliver structured staff training, and establish internal accountability mechanisms.

Market development

FSAs can use data disaggregated by key sociodemographic traits to strengthen market development mandates, including competition and capital markets development. For capital markets development, segmented analysis can help FSAs assess whether certain groups (e.g., women) participate equitably in capital markets as MSME issuers, crowdfunding entrepreneurs, or retail investors. For women-led MSMEs, this may include examining disparities in crowdfunding success rates, average amounts raised, return rates, and investor behavior across gender.

Disaggregated data can also support competition goals by enriching the investigation of various issues, including identifying customer segments that are most vulnerable to declining competition, or analyzing whether FSPs with different levels of gender diversity are more likely to engage in anti-competitive practices. Disaggregated data can be used to identify general levels of competition (e.g., total number of FSPs in each market), market concentration (e.g., number of contracts by FSP in each market), market domination via pricing and other practices, and customer mobility and choice (e.g., successful loan portability requests and data-sharing via open finance).

Open banking and open finance frameworks introduce a further dimension to analyses of competition. Where granular open finance data are collected by FSAs (Tier 3), disparities in data-sharing consent rates and in API performance (e.g., API availability, response time, rejected API calls) by FSP type could signal that open finance benefits are not being realized equitably across FSPs serving vulnerable customer segments. For example, FSAs can encourage industry bodies to monitor the open finance ecosystem and publish performance indicators of participating FSPs, while using regulation to require lenders to enable loan and account portability through open finance, as Brazil has done (Dias et al. 2026).

4. Building Blocks for Disaggregated Data and Segmented Analysis

For most FSAs, moving along the Customer Visibility Tiers aiming at Customer 360 (Tier 3) often entails costs and significant responsibilities, requiring deliberate reforms to strengthen the data collection systems and analytical tools and skills. A data strategy and a robust data governance framework are the basis for gradual, proportional reforms—strictly driven by analytical needs and evolving capacity at both FSAs and FSPs—that protect privacy and ensure ethical data usage. Technology, particularly SupTech, functions as a cross-cutting enabler to support consistent progress (Figure 3).

Source: Authors

4.1 Data Strategy

Many FSAs have progressively amended their regulatory reporting regimes over the years, with a growing number adopting formal data strategies (Cambridge Suptech Lab 2023). As their mandates and data needs expand to reflect market developments, including the large-scale adoption of AI by regulated firms and to power SupTech tools, many FSAs are reforming their reporting regimes. To better structure such reforms, a number of FSAs are adopting overarching data strategies where data are treated as a strategic asset for research, policy, supervision, and regulation.

A data strategy helps FSAs avoid common shortcomings of loosely planned data reforms. Effective planning is important to avoid setbacks such as unplanned costs, and performance and data quality issues. Typically, these setbacks derive from institutional challenges (e.g., lack of buy-in and political support, poor reporting guidance and data standards, duplicative reporting requirements, budget and capacity constraints) and technical issues (e.g., lack of privacy-preserving techniques, outdated legacy systems, manual procedures, lack of audit trails). A longer planning process may yield better results than rushed implementation, and reforms may be most effective when the data strategy is widely consulted with key parties such as reporting FSPs.

A data strategy is a blueprint for FSAs to improve regulatory data and enable the transition to Visibility Tier 3 (see Section 2.1). A data strategy details how the FSA will change its data collection, management, use, sharing and public dissemination, and supporting IT and data infrastructure, based on a detailed diagnostic of the gap between the current state and the desired state. The objectives may include reducing reporting costs for FSPs (e.g., by automating the collection mechanisms and eliminating duplicative reports), strengthening data security, linking to external databases, and enabling the adoption of AI-powered SupTech solutions.⁵

Data strategies typically include:

• Reforms to the reporting requirements: shifting from quarterly aggregated data to more granular and frequent data collection, such as was implemented by the Bank of Thailand (Chansang 2019); adding sociodemographic variables to current data (e.g., NBR, CMF, SFC); developing digital regulatory reporting (e.g., the UK's Financial Conduct Authority (FCA) and Bank of England (BoE)).
• Reforms to the reporting/data collection mechanisms: shifting away from a traditional template-centric system that enable Visibility Tier 3 requires overcoming fragmented legacy systems and fully automating data reporting and collection processes. Novel methods, such as on-demand data viewing using Application Programming Interfaces (APIs) (Restoy 2025), are also becoming more common (e.g., Bank of Uganda's approach to mobile money data).
• Reforms to the data-sharing framework: to break internal silos (e.g., Bank of Portugal's Integrated Data Management, Bundesbank's Integrated Micro-data Based Information and Analysis System (IMIDIAS)) or inter-agency silos (e.g., between OeNB and the Austrian Financial Market Authority (FMA), the European Banking Authority (EBA) and the European Central Bank (ECB), the Bank of England (BoE) and the UK's Financial Conduct Authority (FCA), the Central Bank of Colombia (BanRep) and the Financial Superintendency of Colombia (SFC)).
• Reforms to the IT and data infrastructure/architecture: to enable the use of cloud services, data lakes, graphs, and other techniques that facilitate the combination of structured and unstructured formats.

Table 3. Comparing Aggregated Segmented Data with Granular Data

Source: Authors

When designed with clear standards and phased and proportional implementation, granular data collection tends to yield benefits that outweigh costs over time (Table 3). Incorporating sociodemographic attributes should be embedded in a broader data strategy to ensure quality and contain costs. For FSAs that already collect some datasets in granular form (Tier 2 or 3)—as many do for credit operations—integrating sociodemographic dimensions adds only marginal costs for FSPs and FSAs. FSAs reliant on template-based reporting of aggregated indicators (Tier 0 or 1) may start more selectively, prioritizing a single sociodemographic dimension, such as gender, or limiting initial scope to one or two priority datasets.

4.2 Data Governance Framework

Data governance refers to how an FSA manages and protects data throughout its life cycle, from collection to disposal, ensuring quality, security, integrity, privacy, availability, and usability. A formalized data governance framework is essential to systematize how data are managed, and to ensure compliance with data protection and bank secrecy laws, and internal policies, while increasing transparency and accountability. It encompasses processes, organizational structures, and roles—including those of data owners (who hold decision-making authority over datasets) and data stewards (who carry out day-to-day management)—along with principles and rules for data access, analysis, quality assurance, and sharing.

The data governance framework should assign clear responsibilities throughout the data management cycle, including for sociodemographic variables and personal data. These should include: profiling (examining datasets to identify inconsistencies, missing values, duplicate records, and anomalies); cleansing (correcting, filling, or removing erroneous records); validation (ensuring data conforms to preset rules); metadata management (maintaining information about data origin, common definitions and a shared dictionary, permissible values, and version history); data lineage (documenting the origin of the data and the transformations it undergoes); traceability (maintaining a record of how information is used and modified); change management (governing updates to data models); data integration (including connecting internal and external data sources via APIs); and user profiles and access controls (safeguarding data security and compliance).

As FSAs move towards Visibility Tiers 2 and 3, with growing volumes of sensitive data, robust data governance becomes a necessity rather than just good practice. Regulatory data are largely confidential and subject to data protection, privacy, and bank secrecy laws. Moreover, under some privacy frameworks, gender and other attributes are classified as "special-category" data, subject to stricter processing conditions (Table 4), including Data Protection Impact Assessments (DPIAs).

Table 4. Potential Levels of Privacy Sensitivity of Regulatory Data

Source: Authors

Research has shown that granular data pose reidentification risks even after explicit identifiers are removed (see, for instance, Carvalho et al. 2025, Xia et al. 2023, and ABS 2021). Gender is among the easiest attributes to infer by using ancillary information, and among the most exploitable for targeted harm such as restriction of individual freedoms and gender-based violence. Data governance frameworks must, therefore, embed privacy and data security throughout the full data life cycle.

In addition to ensuring data privacy, FSAs moving along the visibility ladder may embed the following complementary principles in their data governance frameworks:

1) Rationalization and minimization of data collection. This principle requires limiting the collection of data, particularly sociodemographically disaggregated data, to what is strictly necessary (minimization) and adopting collection mechanisms that reduce complexity for FSPs (rationalization). In many FSAs, financial inclusion data reporting is duplicative or inconsistent with statistical and supervisory collections. Addressing this problem requires integrating sociodemographics into mainstream regulatory reporting, using common cross-departmental data standards to ensure interoperability, consistency, and comparability.

2) Maximization of value through sharing and minimization of duplication. Different FSAs and other public authorities may consider integrating their data needs, reporting, and collection frameworks. This entails harmonizing data standards and integrating systems, supported by a clear cross-agency framework for data access and sharing that preserves privacy and security (see Inter-Agency Data Sharing).

3) Transparency and accountability. Data governance frameworks should be formalized and kept current to ensure transparency regarding the collection and use of regulatory data, particularly when sociodemographics are collected in granular form. Data-sharing arrangements with external parties should be addressed, with roles and responsibilities clearly assigned.

4) Responsible and ethical segmented analysis. Data governance frameworks must ensure the ethical use of disaggregated data following a "do-no-harm" principle, especially considering the risks posed by reidentification of granular data. FSAs may start using disaggregated data primarily for diagnostics, testing the explanatory power of sociodemographic variables before jumping into action. Acting on the evidence generated by disaggregated data should never be automated but rather always subject to careful discussion.

4.3 High-Quality Disaggregated Data

When implementing data reforms, FSAs can take targeted steps to improve the quality of disaggregated data—its relevance, granularity, accuracy (credibility), timeliness, accessibility, coherence (comparability), and integrity—across four areas: standardization, validation of sociodemographic variables, data scope and coverage, and external database linkage (Figure 4).

Source: Authors

Standardized Data Frameworks

Data standardization is critical in all Visibility Tiers, but it becomes considerably more demanding as FSAs increase data granularity. Effective data standardization requires providing FSPs with key documentation, designed in consultation with the industry and aligned with applicable international standards, guidance, and definitions issued by standard setters. These include: (1) a data dictionary with uniform definitions for every reported attribute; (2) a data taxonomy or technical data model specifying which attributes to report, in what structure, and using which codes; (3) standardized reporting templates or alternative data transmission mechanisms; (4) transformation rules to convert raw business data into standardized reportable form; (5) uniform validation rules covering format, domain, consistency, and referential integrity; and (6) common identifiers for individuals and legal entities.

When introducing sociodemographic variables, FSAs should consult FSPs early and extensively, plan for implementation challenges—especially when transitioning from aggregated to granular reporting (Tier 1 to Tier 2)—and consider testing periods before full rollout.

Quality of the Gender Variable

To report the gender variable, FSPs must first add the relevant data field to their systems, using visual assessment, ID-based verification, or self-reported information (Figure 5).

Source: Authors

FSAs may consider requiring FSPs to validate gender (and any other sociodemographic data) against a national ID or other official document or databases, regardless of the collection method initially used by FSPs. Where legal frameworks recognize gender identities beyond the binary, FSAs may consider aligning their data collection accordingly, with the objective of capturing the financial sector experiences of gender minorities (BIS 2022b, SFC n.d.).

To support a range of national policies and FSA-specific measures to develop the MSME sector, FSAs could consider improving gender-disaggregated data related to MSMEs. Gender for MSMEs is typically assigned by identifying the gender or sex of the business owner or main executive. Despite receiving growing attention, gender-disaggregated MSME data remain underdeveloped (AFI 2026, We-Fi 2025b). Practical challenges include inconsistent definitions across agencies, business registries that lack reliable ownership and leadership data, and the higher likelihood that women-led nano and micro-enterprises operate informally and fall outside official registries.

Data Scope and Coverage

Data scope defines the content and boundaries of required reporting. Two dimensions arise for operational data: the range of product types and accounts reported; and the transactional activity within those products. Beyond product and transaction coverage, scope also determines whose sociodemographic data are collected, which does not need to be limited to customers: it can include investors, loan guarantors, and FSP staff. Data coverage refers to the range of FSPs required to report. As noted in Section 2.1, FSAs typically begin granular data collection in the domains where reporting frameworks are most developed (usually credit) before expanding incrementally up the Visibility Tiers. Any expansion of regulatory disaggregated data scope and coverage must respect the rationalization and minimization principle (Section 4.2) and be carefully planned as part of a broader data strategy.

Linkages to External Databases

Where national ID systems and business registration systems exist, FSAs and FSPs can leverage them to collect and/or verify gender and other sociodemographic information, using the ID or registration number as the linking identifier. National ID systems typically store many personal attributes such as name, date of birth, and sex (and/or gender), and may also include biometrics. Accessing these databases can reduce inconsistencies between self-reported data and official records, promote harmonized standards, reduce frictions in FSP-customer interactions, and support KYC compliance. However, these databases, particularly in EMDEs, may present practical shortcomings which need to be carefully considered (Figure 6).

Source: Authors

Access to additional external databases—statistical offices, tax authorities, labor departments, gender monitoring agencies, and credit registries held by other authorities—can further enrich segmented analysis. The Central Bank of Brazil illustrates a Tier 3 approach where linkages to some external databases is enabled: rather than requiring FSPs to report sociodemographics, it collects tax numbers in granular datasets and uses them to retrieve gender and other attributes from the tax authority, labor department, ministry of justice, and ministry for social development, under formal and permanent data-sharing arrangements.

4.4 Capacity for Segmented Analysis

To effectively leverage disaggregated data using sociodemographic variables, FSAs may need to strengthen their analytical capacity in alignment with institutional priorities. The ability to do segmented analysis and uncover insights hidden in aggregate indicators does not automatically follow improved data disaggregation. Advanced analytical capacity is needed to manage growing data volumes and equip supervisory and regulatory staff with the skills to interpret sociodemographic data, assess materiality, and integrate findings into supervisory and regulatory decisions.

A common barrier is the perception that sociodemographic data are peripheral to core institutional mandates, with gender and other variables often viewed as irrelevant to prudential or conduct supervision. Practical approaches to overcome this include: framing sociodemographically segmented analysis as an enhancement of existing objectives; using pilot projects and peer learning to demonstrate value; involving supervisory staff early in the design of data reforms; and providing clear technical guidance, including concrete use cases for integrating disaggregated data into routine supervision and regulatory analysis.

4.5 Inter-Agency Data Sharing

Implementing the data governance principles laid out earlier in this paper may require FSAs to establish or adapt existing governance and technical mechanisms for the safe sharing of disaggregated regulatory data with other agencies. Advocating the principle of "define once, report once, use many times", the Irving Fisher Committee on Central Banks Statistics (IFC-BIS) highlights that enhanced inter-agency data sharing improves quality, promotes standardization, reduces duplication, and lowers compliance costs for reporting institutions (IFC-BIS 2017, 2024, 2025). In practice, this involves two interrelated dimensions:

• Data sharing arrangements: transition from ad-hoc data exchanges to institutionalized frameworks supported by instruments such as Memoranda of Understanding (MoUs) to define limited sharing purposes, permitted uses, access rights and controls, privacy-protecting techniques, and strict security safeguards (IFC-BIS 2024).
• Integrated reporting mechanisms: integrate requirements across authorities to reduce duplication and systematically embed sociodemographic attributes into reporting, improving data quality and lowering FSP compliance costs (IFC-BIS 2025).

Despite clear benefits, FSAs face multiple challenges in establishing effective data-sharing and reporting integration arrangements. Legal constraints related to data protection, confidentiality, and banking secrecy can restrict the sharing of granular data (IFC-BIS 2024), highlighting the importance of early legal assessments to distinguish perceived constraints from actual prohibitions. Operational barriers, including incompatible IT systems, lack of common identifiers, and resource constraints, alongside cultural and organizational factors such as limited trust, can further impede cooperation.

4.6 Dissemination of Data as a Public Good

Alongside inter-agency arrangements, FSAs can create further value by strengthening public dissemination of nonconfidential, disaggregated regulatory data and related analyses. There is a growing case for disclosure of segmented metrics, with gender-related metrics representing the most developed and immediately actionable starting point (BIS 2022b). Treating data as a public good strengthens transparency and accountability, and enables a range of actors, including investors, development partners, advocates, researchers, and FSPs to make better-informed decisions (Table 5).

Table 5. How Dissemination of Disaggregated Regulatory Data Creates Value for the Wider Market

Source: Authors

FSAs can disseminate disaggregated data through a variety of channels tailored to different audiences and their own internal capacities. Common approaches include public-facing interactive or static dashboards, such as the Bangladesh Women's Financial Inclusion Data Dashboard, the Nigerian Women's Financial Inclusion Dashboard, the Central Bank of the Philippines' Gender and Development Corner, and NBR's Financial Inclusion Dashboard. Another approach is periodic reports, such as the Chilean CMF's Gender in the Financial System reports, the Central Bank of Brazil's Risk Indebtedness and Financial Citizenship reports, and the US CFPB's Mortgage Market Activity and Trends reports. Some FSAs complement these communication instruments with controlled access mechanisms, i.e., secured research environments that allow approved users to access granular data with controlled outputs.

4.7 Technology as a Cross-Cutting Enabler

Advances in SupTech and regulatory technology (RegTech) are facilitating both FSAs and FSPs to move toward more data disaggregation and granular data reporting. These technologies support the full regulatory data life cycle—collection, validation, management, analysis, visualization, and dissemination—while improving data quality, efficiency, and policy relevance. Technology is a cross-cutting enabler, impacting all building blocks and therefore what FSAs can plan for in their data strategies.

SupTech tools can strengthen FSAs' data collection, storage, management, and usage capabilities, while RegTech solutions can help FSPs implement new reporting requirements and improve the data quality of existing requirements. Cloud-based infrastructures, data lakes, and centralized or federated data repositories provide FSAs with the capacity to store and manage large volumes of granular data. Machine learning (ML) techniques can be used by FSAs for anomaly detection, outlier identification, and prioritization of data issues across large granular datasets.

AI, ML, and natural language processing tools enable segmentation, trend detection, and behavioral analysis across sociodemographic groups. LLMs extend these capabilities further, enabling the automated analysis of vast volumes of unstructured text. Tech-powered visualization tools can more easily translate segmented data into powerful communication for actionable insights. Tools such as interactive dashboards and geospatial mapping allow FSAs to integrate sociodemographic segmentation into internal routine monitoring and decision-making. Despite its benefits, technology can also create risks for FSAs, which need to be managed. When automating the analysis of disaggregated data using AI and ML models, FSAs should ensure that training data are inclusive and subject to human oversight.

5. The Implementation Pathway

Sections 3 and 4 set out the analytical case and institutional building blocks for embedding sociodemographic disaggregation into regulatory data flows and FSA decision-making. Translating this framework into practice requires each FSA to make choices based on where it currently stands and where it wants to go, whether that means Tier 3, with Customer 360 Visibility and linkages to external databases, or another point on the customer visibility ladder. This section addresses key implementation considerations (Figure 7), including an initial self-assessment to identify the FSA's current Visibility Tier, the legal and governance foundations for reform, data strategy design considerations, and monitoring and evaluation needs.

Source: Authors

5.1 Identifying Your Visibility Tier

A candid self-assessment of the FSA's current customer visibility capability is an essential first step of any data reform plan. To design their data strategies (see Section 5.3), FSAs need a good understanding of what their current data infrastructure enables and where the main binding constraints lie. The self-assessment questions in Figure 8 provide an entry point, and Annex 2 provides a more detailed self-assessment questionnaire. These are designed to identify the highest Visibility Tier (Section 2.1) at which an FSA can currently conduct segmented analysis, based on what data it already holds and whether it can link records at the customer level.

Source: Authors

The result of this exercise—including the application of the questionnaire in Annex 2—is not an evaluation of institutional performance or benchmarking, but a simple diagnostic. It tells the FSA where it has analytical footing today and, by extension, where the most valuable next investments lie, relative to its own analytical priorities and its desired state. Two important qualifications apply when identifying the current Visibility Tier:

• First, the Visibility Tier assessment is dataset-specific, not institution-wide. An FSA may operate at Tier 2 or 3 for credit data while remaining at Tier 0 for deposit or insurance data. Data strategies, including detailed implementation plans, should therefore consider the different realities of different datasets.
• Second, there is often a gap between formal data availability and actual data usability. An FSA may receive granular data containing sociodemographic variables, but if those fields are sparsely populated, inconsistently coded, or rarely validated, the effective Visibility Tier for sociodemographic analysis may be lower than the formal reporting architecture suggests.

5.2 Ensuring Legal and Governance Foundations

A formalized data governance framework (see Section 4.2) and a clear legal basis for collecting, using, and sharing sociodemographically disaggregated data are key foundations for the reforms needed to advance across the Visibility Tiers. Considering that legal and regulatory constraints can stall the implementation of data strategies and improvements to the data governance framework, a readiness assessment of the legal framework will allow FSAs to identify real constraints and distinguish them from perceived ones. Such a readiness assessment could cover aspects such as:

1) Legal authority for sociodemographic data collection. FSAs may need to confirm the legal basis for collecting sociodemographic variables as part of mandatory regulatory reporting. In many jurisdictions, this authority exists under broad supervisory or oversight mandates, but where gender or other sociodemographic information is classified as "special category" under data protection laws (see Section 4.2), explicit legal authorization may be required.

2) Data protection and privacy compliance. In some jurisdictions, a DPIA may be required or advisable before processing sociodemographic data at scale and in granular format.

3) Banking secrecy and data confidentiality rules. FSAs may need to assess whether planned inter-agency data-sharing arrangements are permissible under bank secrecy rules and other legal provisions. Constraints that appear absolute may be addressable through properly designed technical and governance arrangements and safeguards such as privacy-enhancing techniques (see Box 8).

5.3 Designing the Data Strategy

The data strategy (see Section 4.1) provides the practical framework for coherent, cost-effective investment where sociodemographic disaggregation is treated as an integral component of a broader data reform—not a parallel track. The following general principles can help FSAs develop a solid strategy for transitioning toward higher Visibility Tiers:

1) Adopt a staged implementation approach starting where data quality is strongest and analytical need is clearest. Experiences from Rwanda and other countries show that the most lasting reforms build on existing high-quality data within a product domain rather than simultaneous improvement attempts across multiple products. Credit data, where stricter regulatory controls typically apply, tends to offer the strongest foundation for starting data reforms.

2) Align improvements in data quality with analytical priorities. FSAs should take a phased approach to improving data quality. A fundamental first step is greater standardization through common data dictionaries and validation rules. The expansion of data attributes, collection scope, and coverage should follow demonstrated capacity in initial domains and be fully aligned with analytical priorities.

3) Build toward inter-agency data sharing and public dissemination. Inter-agency data-sharing (Section 4.5) and public dissemination (Section 4.6) should be built into the data strategy from the outset, but they depend on a solid foundation of data quality and governance.

4) Embed investments in analytical and broader institutional capacity. Strengthening analytical capacity (Section 4.4) and broader institutional capacity must be integral to the data strategy, with capacity-building efforts beginning before new data flows start.

5) Anchor the data strategy to other policy priorities to ensure continuity. Institutional continuity is essential for reforms that span multiple years. Embedding the data strategy into formal institutional documents, such as multi-year supervisory plans, or positioning it as a precondition or strategic lever for other policies (e.g., as the basis for an NFIS monitoring framework) provides protection against discontinuity.

6) Set a realistic financial and cost management plan with staged technological and institutional investments. As a fundamental enabler of data infrastructure and analytical capabilities, technology investment (Section 4.7) should be fully integrated into the data strategy and aligned with the overall phased and proportional approach. Table 6 summarizes what each tier transition requires.

Table 6. Moving Up the Customer Visibility Tiers: Requirements and Cost Considerations

Source: Authors

5.4 Ensuring Proportionality of a Phased Approach

To effectively implement a phased progression across the Visibility Tiers, FSAs should ensure the overall proportionality of their data strategy. As noted throughout this paper, new data collection should be justified by data analytical needs strictly anchored in, and proportional to, the FSA's mandates and the mandates of other authorities the FSAs will share the data with. Likewise, data reforms should be commensurate with the FSA's capacity to use increasingly granular data while protecting privacy. Moreover, proportionality should inform how FSAs adjust and phase the expansion of data collection, carefully considering the different levels of preparedness among FSPs. FSAs may consider staging the imposition of expanded data disaggregation requirements or linkages with external databases, starting with large entities with greater capacity, while giving smaller entities more time to make the necessary investments.

5.5 Adopting a Collaborative Approach

It is important that FSAs adopt a collaborative approach and engage in early consultation with other authorities and FSPs to implement their data strategy. New data reporting standards and validation rules must be extensively discussed with FSPs and partner authorities, and pilots can be useful to both minimize quality issues and help FSAs achieve proportionality. Engagement planning should encompass at least the following stakeholders:

1) Regulated financial service providers. FSPs bear the reporting burden and are central to data quality, making early engagement—ideally during data strategy development—essential to identify feasibility constraints, align analytical objectives, and set realistic timelines.

2) Staff at FSAs. Internal resistance can arise when reforms challenge established routines. Addressing these issues requires a consistent, leadership-endorsed narrative linking segmented analysis to institutional mandates, supported by practical use cases. Change management—through pilot projects, cross-departmental working groups, and internal showcases—can help build ownership.

3) Other public authorities. These can be essential partners, especially when the data strategy includes inter-agency data-sharing arrangements with other FSAs, consumer protection agencies, or competition agencies. Also, implementation may require coordination with statistical offices for benchmarking and survey data, and with ID and civil registries for data linkages underpinning Tier 2 and Tier 3 analyses.

4) International engagement. FSAs can draw upon the international community when designing and implementing their data strategy. Global and regional forums—such as AFI, the IFC-BIS, the World Bank, the WE Finance Code, the Alliance for Innovative Regulation (AIR), CGAP, and the Cambridge Centre for Alternative Finance—provide platforms to exchange knowledge about granular data frameworks, governance, SupTech, and data modernization.

5) Academic and research institutions. These institutions can strengthen an FSA's capacity to implement this guide by bringing methodological expertise and analytical bandwidth that FSAs may lack, helping them design sound disaggregation frameworks, validate findings, and translate raw data into evidence.

5.6 Monitoring and Evaluating Progress

Sustaining reform over time requires clear milestones, regular progress reviews, and feedback mechanisms that connect analytical outputs with ongoing efforts to improve data quality. FSAs can structure implementation monitoring around a periodic review covering all implementation items, from foundations to the outputs devised in the data strategy. A solid monitoring and evaluation framework would identify milestones and related progress indicators or markers for each aspect of the data strategy. A progress-tracking dashboard can help operationalize such monitoring, as well as internal and external reporting. It could include key markers for each implementation aspect, such as:

1. Analytical needs assessment: completion of a baseline assessment of analytical needs for disaggregated data, per an institutional mandate, and the related prioritization of datasets and breakdowns, by type of regulated FSP.
2. Governance and legal framework: completion of the legal readiness assessment, resolution or mitigation of legal gaps; drafting and approval of the data governance framework; and establishment of required inter-agency governance arrangements.
3. Stakeholder engagement: completion of a stakeholder mapping exercise with an engagement strategy for each key stakeholder; implementation of prioritized engagement strategies.
4. Technology and infrastructure investments: completion of a diagnostic of current IT and data infrastructure gaps against the target Visibility Tier; assessment of the need for external vendors versus internal capacity; upgrade of the central data repository to support granular data storage; and deployment of cross-departmental data-sharing infrastructure.
5. Data quality: imposition of new reporting requirements for datasets prioritized for sociodemographic disaggregation; adoption of standardized data dictionaries and validation rules; completion of the first reporting cycle; completion of a quality review; and establishment of a regular data quality monitoring process.
6. Internal analytical capacity: completion of a sensitization and capacity building plan; integration of disaggregated evidence into at least one supervisory activity under a prioritized mandate; completion of a first-cycle training program; and establishment of a cross-departmental working group on ethical segmented analysis.
7. Public dissemination: publication of a first set of disaggregated indicators through the FSA's existing dissemination channels (e.g., annual report); design and implementation of a controlled data access environment for researchers.

Conclusion

This Guide offers a conceptual framework, including frontier use cases, institutional building blocks, and key considerations for the implementation of reforms for FSAs to embed disaggregated analysis into their decision-making in the exercise of their policy, regulatory and supervisory responsibilities. Its central argument is that sociodemographic disaggregation should shift from being a peripheral enhancement to being recognized as an additional means to contribute to FSA mandates such as financial inclusion, consumer protection, and prudential goals.

The use cases described in Section 3 illustrate what becomes possible when authorities move beyond aggregates: identifying gaps in access and usage among underserved populations, refining risk assessments, potentially adjusting regulatory requirements, and detecting market conduct and competition weaknesses that aggregate figures would otherwise obscure. The building blocks and implementation pathway set out in Sections 4 and 5 are designed to make this transition actionable. Section 4 explains the building blocks, starting with the need to establish robust data governance and a data strategy. Key considerations for a phased, proportional and collaborative implementation of the building blocks are presented in Section 5, starting with a self-assessment of the current Visibility Tier, combined with the identification of the analytical questions that matter most.

Progress is incremental, and each step compounds over time. With an implementation approach aligned with existing capacities and priorities, alongside systematic monitoring and evaluation, FSAs will be better equipped to manage the risks that come with these complex reforms, including internal and external resistance, discontinuity of financial commitment, and technical challenges such as the need to protect privacy when handling granular data.

A call to action. FSAs are encouraged to embed sociodemographically disaggregated data into their core functions, starting with diagnostics that help identify gaps, risks, and distributional impacts, and gradually moving toward acting upon the evidence generated by such data, where their mandates and institutional context warrant it. In doing so, FSAs will align with a broader global shift toward more granular and inclusive financial data and analysis. This will require that they embed sociodemographic disaggregation into data governance frameworks and data strategies that protect privacy, build the FSA's capacity to interpret and act on disaggregated data, and engage FSPs, peer authorities, and other relevant stakeholders as partners throughout the process in a phased, proportional, and highly collaborative approach. Even FSAs currently at Tier 0 of Customer Visibility can start taking meaningful steps to put in place the necessary conditions to gradually climb up the visibility ladder. The guidance provided here, together with the CGAP LENS tool, offers a structured basis for doing so.

Footnotes

1. CGAP LENS is available at https://www.cgap.org/lens.
2. FSAs do not need to link or integrate all their granular data. The use of random samples to create representative extracts or panels of individuals can satisfy many analytical use cases, and be used to construct indicators with sufficient precision for monitoring and comparing change over time or between population segments.
3. CGAP LENS is freely available at https://www.cgap.org/lens.
4. These five dimensions are used strictly as an organizing device for the analysis proposed. Several global organizations have already issued definitions of financial inclusion dimensions that differ in number and scope (e.g., AFI 2019; GPFI n.d.; World Bank 2012), and this Guide does not seek to redefine terms.
5. Data strategies are closely linked to SupTech strategies. They help FSAs secure and transparently allocate funding for multi-year reforms to IT and data infrastructure, and incentivize other actions to enable the implementation of the data strategy.
6. The WE Finance Code is explicitly built around the premise that gender (sex)-disaggregated MSME finance data is systematically lacking and urgently needed. Its core data pillar commits FSPs to collecting and reporting five core sex-disaggregated data points on their MSME portfolios.
7. To help address this challenge, the Alliance for Innovative Regulation (AIR) applies their Innovation Elements Framework to help FSAs understand and prioritize the organizational capabilities that underpin the mainstreaming of gender disaggregated data in their institution and financial system (https://regulationinnovation.org/gender-disaggregated-data-for-inclusive-regulation/).