Ethiopia’s Next Financial Inclusion Strategy Must Be Geo-Enabled, Data-Driven, and People-Centered

Ethiopia’s National Financial Inclusion Strategy (NFIS II) helped unlock rapid growth in digital finance. But accelerating access has revealed a deeper challenge: inclusion is not yet translating into meaningful use for millions. As the country prepares NFIS III, now is the moment to shift from counting access points to ensuring financial services actually work for everyone, everywhere.

This article is an output of AKOFADA (Advancing Knowledge on Financial Accessibility and DFS Adoption), a project working to increase knowledge and transparency within Ethiopia’s DFS ecosystem. 

Ethiopia’s National Financial Inclusion Strategy (2021-2025) (NFIS II), now nearing the end of its period, has laid a strong foundation for the expansion of digital financial services. Since NFIS II reforms began, digital finance adoption has accelerated: debit cards and mobile banking grew strongly, and mobile-money accounts have quadrupled. This momentum was supported by key regulatory reforms like the payment issuer directive in 2020, which opened the door for non-financial fintechs and mobile money operators. Consumer protection has also made meaningful strides: the Financial Consumer Protection Directive (FCP/01/2020) set legally binding obligations for fair treatment, clear disclosures, and grievance redress mechanisms across financial providers, helping build trust in the system.

Yet three key structural gaps limit impact. First, a targeting gap: NFIS II reports aggregate metrics like the percentage of adults within 5 km of an access point, but does not require service points to be geotagged or mapped at the kebele level. Without a live national access map, precision targeting in underserved areas remains difficult.

Second, while the strategy acknowledges disparities across gender and geography, it stops short of assigning binding sub-targets or tailored delivery approaches for vulnerable groups. These include pastoralist communities in Afar and Somali, communal societies in SNNP or Gambella, people with disabilities, internally displaced persons (IDPs), and Muslim-majority areas needing Sharia-compliant products. Their distinct financial needs are largely unaddressed in implementation.

Lastly, there is an execution gap where the strategy tracks static counts, such as the number of accounts or access points, but not active usage, service reliability, or user experience. Key metrics like monthly active users, agent liquidity and uptime, transaction failures, or time-to-restore after outages are missing. World Bank assessments reinforce this concern: despite infrastructure expansion, many rural communities remain excluded due to weak targeting, limited usability, and insufficient follow-up. Together, these gaps suggest that while access has expanded on paper, inclusion in practice still lags, calling for a sharper focus on geospatial precision, inclusive design, and performance-based measurement.

It is partly because of these policy gaps that the 2025 Global Findex came with only a slight increase in account ownership in Ethiopia, rising from 46% in 2021 to 49% in 2024, far short of the transformative progress many had hoped for.

As Ethiopia prepares to launch its next financial inclusion strategy (NFIS III), it is essential to address existing gaps and ensure that the next five years are guided by an improved approach.

Map-First Inclusion and a National Financial Data Commons

The first step toward smarter financial inclusion is making the invisible visible through a map-first approach that overlays financial access points (Banks, ATMs, POS machines, Micro Finance Institutions (MFIs), Saving Credit Cooperatives Organizations (SACCOs), and mobile-money agents) with infrastructures (telecom and electricity), population data, roads, and socioeconomic indicators.

This creates a live national geo-diagnostic that shows where digital finance can scale immediately (when people, signal, and power align), where bundled solutions are needed (for example, signal without power or agents), and where service is truly absent and why.

Ethiopia’s financial inclusion data remains fragmented across institutions, with no live national map or requirement to geotag branches and mobile money agents. Most locations are recorded only at broad administrative levels rather than precise GPS points, making it difficult for policymakers to identify service gaps or target interventions effectively.

A National Financial Data Commons could address this challenge by aggregating de-identified, geotagged data from banks, microfinance institutions, telecom operators, the Ethiopian Electric Utility, the Ethiopian Electric Power, the Ethiopian Geospatial Information Institute, and the Central Statistics Agency base maps. The result would be a near-real-time, interactive map showing access points, patterns, as well as financial deserts. This shared system would enable regulators and providers to pinpoint underserved areas and allocate resources more strategically.

Importantly, Ethiopia is not one market. A geo-enabled NFIS built on the Commons must adapt to regional mobility, infrastructure, and belief systems. Pastoralist regions like Afar and Somali need mobile agents, offline wallets, and even livestock-linked financial services. Communal and religious contexts require belief-aligned products, assisted onboarding, and local-language outreach. Overlays with roads, schools, and poverty maps can sharpen delivery, helping target high-impact gaps and low-cost wins.

The Commons should also publish a few core metrics that drive action, such as the share of adults within 5 km of a functioning access point, monthly active use by region and segment, agent uptime and liquidity, failed or reversed transactions, and time to restore after outages. Privacy would be preserved by design, and only aggregated data would enter the system, protected through k-anonymity, data minimization, and differential privacy where needed, under multi-stakeholder governance and audited APIs.

Together, a map-first approach and a National Financial Data Commons would turn inclusion management from static reports into a living, data-driven system. This combination enables policymakers and financial institutions to design tailored, efficient, and inclusive financial strategies, ensuring the right tools reach the right places at the right time.

The Data Commons can also be used beyond mapping access points to reveal usage patterns. Despite growth in bank agents and outlets, uptake and active use remain low, with many agents reporting cash shortages, system outages, and failed transactions. Counting access points isn’t enough. 

Dormant accounts and unreliable agents mask real gaps. To move from access to inclusion, Ethiopia must track quality and resilience: active account use, agent uptime, failure rates, and customer experience (cash-out distance and service continuity). Kenya’s mobile money success and India’s dormant-account challenge both show that real inclusion requires going beyond numbers to ensure services work and are used.

However, building a financial data hub also requires careful management of data and system risks. Aggregating data and using privacy-preserving techniques can help protect individuals. For instance, differential privacy methods introduce statistical “noise” so that aggregated reports reveal usage patterns without exposing personal details. Strict access controls and regular audits will further strengthen data protection. 

Experience in Ethiopia shows that providers may overreport or make errors. To address this, the system should include validation rules such as balance checks and plausibility limits, along with scorecards or feedback loops to flag and correct problems. The reporting process should also be designed to reduce the burden on agents and firms by collecting only core indicators in a lean schema and automating data submission as much as possible through APIs.

Accessibility-by-Design (What NFIS II Misses)

People with visual impairments in Ethiopia still face major barriers using ATMs, USSD, and mobile apps. Most ATMs are not voice-assisted, and many banking apps lack proper screen-reader support, forcing users to rely on sighted help and compromising their independence. 

Yet, NFIS II includes no binding standards, making accessibility optional and uneven. To close the gap, Ethiopia should mandate toll-free IVR (Interactive Voice Response) services in major local languages; simplified, screen-reader-friendly USSD flows; WCAG 2.1 AA (Web Content Accessibility Guidelines, Version 2.1, Level AA) compliance for mobile apps with annual audits; and physically accessible ATMs (audio jack, tactile keypads, ramps).

WCAG 2.1 Level AA are international standards from the World Wide Web Consortium (W3C) for making web content and mobile apps accessible to people with disabilities, covering a wide range of needs like blindness, low vision, hearing loss, and limited movement. This standard is legally required for state and local governments in the US and the European Union to ensure digital content is understandable, operable, robust, and perceivable. 

The Bank of Abyssinia’s pilot of talking ATMs proves this is locally feasible. These upgrades enhance usability for all while restoring autonomy for vulnerable users. Progress can be tracked via simple metrics like IVR uptake, app audit scores, and % accessible outlets per financial institution.

What Makes This Different from Typical Coverage?

This approach shifts focus from counting accounts to mapping how and where inclusion happens. Instead of just tallying users, it employs geospatial and usage analytics. For example, World Bank guidance recommends mapping financial access points and identifying service gaps using GIS. Overlaying such maps with population density, poverty, or infrastructure (as done in Pakistan’s inclusion strategy) highlights underserved clusters. Likewise, we should prioritize usage and quality metrics (transaction volume, uptime, fees) and accessibility (e.g., disability-friendly interfaces) over mere ownership rates. Data will be updated and published regularly (a live dashboard), not hidden in static reports. 

Indeed, geospatial data can be used as a supervisory tool or even opened as a public good. Finally, unlike passive surveys, this proposal imposes obligations on providers and builds a national data backbone to continually track inclusion. Ethiopia’s NDPS itself underscores this approach by mandating monitoring “using key performance indicators.” In short, by moving from tallies to maps and from isolated data silos to an integrated, real-time platform with enforceable standards, Ethiopia can target and accelerate true financial inclusion.