Challenging the false dichotomy: Education innovations that reach the most vulnerable AND at scale

This blog was written by Hannah Graham, Managing Director of CGA Technologies, based on her presentation at the September 2025 UKFIET conference. 

Despite decades focused on developing inclusive education systems that address the needs of the most marginalised (with many notable gains and improvements), the global population of out-of-school children has increased by 20m to 272m (UNESCO 2023).  

With education aid projected to fall 25% between 2023 and 2027 (Global Education Monitoring Report 2025), some will ask, “can we continue to target the hardest to reach or should attention shift to scale?” 

The answer is neither. Instead, there is a third way: innovation that leverages existing systems and services to reach the most vulnerable – more efficiently, effectively, sustainably and – crucially - doing so at scale. 

At UKFIET 2025 this week (18 September), I introduce the preliminary results from the first year of the Girls In School Initiative (GISI), a unique data-led CashPlus (1) education programme aimed at increasing school attendance and retention rates among vulnerable girls in Sierra Leone.   

Its uniqueness comes from leveraging data from an existing government-owned data system (Wi De Ya - a teacher and learner enrolment and attendance monitoring system developed by CGA Technologies) to identify and target the most vulnerable in school girls as well as formerly out-of-school girls to receive lightly conditional monthly cash payments, delivered via mobile money.  

That same platform, utilised by teachers and government staff, is then used to monitor the girls’ attendance. Those with poor attendance are flagged for follow-up and linked to existing complementary services, such as health, counselling, legal support, and disability or child protection services. 

Our theory being that, by adopting a data-led approach and using robust, highly disaggregated data which continuously flows in from existing government systems, we eliminate the need for standalone needs assessments, large-scale surveys and resource heavy monitoring. By targeting and verifying girls based on data already recorded by school leaders in Wi De Ya – and monitoring them using the same system – GISI is more efficient and cost-effective. 

The impact of GISI on participating girls’ school attendance is encouraging, demonstrating that leveraging existing mainstream systems to target vulnerable learners can be effective without requiring major additional expenditure or parallel, unsustainable structures. 

Over its pilot year (AY 2024/25), GISI, which is funded by Corus International, targeted 853 vulnerable primary school girls and former out-of-school girls throughout the academic year. Preliminary findings (2) drawn from the profiles and twice-daily attendance records of more than 12,000 learners (including those who received cash transfers) indicate a modest but significant improvement in school attendance among participating girls. From January to April, their attendance was 4 percentage points higher than both the 88% average of all learners and the control group.    

Over-age learners were significantly more responsive. Within each age group, the attendance of programme participants was 3.5 percentage points higher than non-participants, except among learners aged 14-16 where the attendance of those receiving cash was 10 percentage points higher than those who didn’t. 

Girls and caregivers reported wider gains, who used the funds helped cover school and basic household needs. Having school materials and uniforms instilled pride in the girls who reported that this motivated them to attend school; and they had money to buy lunch, making them more attentive in class. Caregivers reported using cash for food, electricity, medical expenses and hygiene, supporting all members in the household.  

The beauty of this approach is its scalability – a hypothesis we are testing as GISI enters its second year. This year we are doubling our efforts to target 2,000 vulnerable girls across 60 primary schools (up from 29 in year one) in five districts (up from four).  

As the Education Ministry (MBSSE) and Teaching Service Commission scale Wi De Ya to all primary schools (3) and later all Junior and Secondary schools, so then GISI can scale – as can any programme that partners with the Ministry and adopts this framework.  

Using existing mainstream systems and eliminating the need for standalone needs assessments, large-scale surveys and resource heavy monitoring, is precisely what makes scaling programmes like GISI possible.  

The Global Education Monitoring Report 2023 cautions practitioners that "technology use [in classrooms] is not universal and will not become so any time soon". Therefore, our focus in low resource settings should not be getting children on devices, but empowered teachers, empowered management, and better resourcing. Focus then, should be on developing the processes and the systems – knowing who and where learners are and what their needs are, where teachers are and are not. 

GISI shows what is possible when data systems and partnerships are aligned to deliver greater impact—a critical need as education budgets come under increasing strain. 

Across the sector, siloed systems and disconnected programmes drive duplication and inefficiency. By contrast, governments are investing in core critical digital infrastructure to strengthen their own services. When education partners work through these systems, they not only boost efficiency and sustainability, but also help build national capacity and coherence. Interoperability at the national level is key: this is where resources and data are most effectively aligned with the needs of populations. 

Access to government-owned systems gives all partners a single, trusted source of information, reducing duplicative data management costs and ensuring more funding is directed where it is most needed. Embedding programmes in government systems also strengthens digital public infrastructure, while giving stakeholders—government and project staff alike—the tools to work transparently and with greater control. 

Creating meaningful digital public infrastructure is not about launching new platforms but about connecting existing systems with new solutions in ways that are realistic, sustainable, and rooted in government structures and processes. That requires interoperability, long-term support, and above all collaboration—project teams, government staff, and partners aligning data and processes together from the outset. 

As attention turns to technologies like AI, getting digital public infrastructure right is even more pressing so that the education data feeding into it is robust, accurate, and used effectively and ethically to serve the most vulnerable. 

The building blocks already exist. What is needed now is a collective commitment across the education sector to use them to their fullest, ensuring scarce resources go further and the most vulnerable children continue to receive the support they need—efficiently, effectively, and sustainably. 

Note: In the second phase of the Girls In School Initiative (AY25/26), CGA Technologies will scale the programme to target 2,000 vulnerable girls across five districts in Sierra Leone. GISI is only possible thanks to collaboration with the Ministry of Basic and Senior Secondary Education, which permits us to access and use select Wi De Ya data to deliver the programme. CGA’s sister organisation, LWR, is also implementing the programme in Mali.   

1: Cash Plus interventions combine cash transfers with complementary, context-relevant interventions, in this case intervening to prevent process of school drop out as attendance falls. 

2: Pending fuller analysis.

3: Wi De Ya piloted nationally in 300 government primary schools.

Wi De Ya is the Sierra Leone name for HereMIS, CGA’s education monitoring system. Watch our video to find out more: HereMIS: A data-driven approach to improving education systems