What it actually takes to scale through government: reflections from Oxford during Skoll World Forum 2026

The question of how to move beyond pilots is well established across the sector. Not whether innovations work – many genuinely do – but what happens next. Proven approaches stay small, pilots end, and the cycle starts again.

That challenge was at the heart of many conversations in Oxford last week, including a session where Dr Fay Hodza, PEAS’ Global Senior Director of Programmes, joined implementers and advisors from across health, education, and social services to discuss what it actually takes to move from innovation to national impact.

The conversation started with a question most organisations are still figuring out how to answer: why doesn’t more of this work stick?

Here are the key takeaways from the session:

1. Scaling means working with government, not presenting to it

One of the clearest points to come out of the session was this: scaling through government requires partners to really understand government – not as a delivery partner, but as an institution with its own priorities, budgets, political pressures, and risk calculus.

Ministries often see partnering with NGOs as a risk: political, financial, reputational. If that’s the starting point, the job isn’t to present a model and wait for buy-in. It’s to understand what government is trying to solve, show that your approach helps (with data they can use to justify the decision), be ready to adapt it to work within the system, and reduce the risk they perceive in adopting it.

For PEAS, this shapes how we enter government partnerships. We co-design from the start – working within the system’s constraints and responding to national priorities. The goal is an approach a ministry can own, fund, and sustain. If it can’t survive without us, can we truly say it’s scalable?

2. Practical barriers are easy to underestimate

Operational realities are easy to underestimate from the outside. For example, procurement takes time – often years. Signed contracts don’t guarantee timely funding. Budget cycles don’t align with programme timelines. Funders who want to see government-embedded, sustainable impact need to understand that this requires working capital support during the transition period. The system does not move at the speed of a pilot.

A recurring theme in the discussion was the importance of defining roles early. If you haven’t clarified from the beginning who funds what, who runs what, and who is accountable, it won’t stick. Assumptions that feel obvious at the design stage can become points of failure at implementation.

3. Advocates at all levels are important

The temptation can be to build relationships at the highest level – ministers, permanent secretaries, senior officials – and assume that agreement at the top will translate into action below.

Real traction comes from finding advocates within government at multiple levels. Identifying the department whose mandate most closely aligns with your work, understanding who holds the budget authority, bringing them in early, and making them part of shaping the approach. If they understand and believe in the approach, they will champion it to other departments more effectively than any external organisation can.

This mirrors our own experience. In Uganda and Zambia, we’ve worked to build relationships across the Ministry of Education – not just with senior leadership, but with district officials, school inspectors, and the civil servants responsible for day-to-day delivery. Government is made up of people, and change happens through them.

4. The model needs to be grounded in reality

There is a tendency in the sector to design programmes for ideal conditions, which often don’t exist.

Scaling through government, particularly in low-resource and rural contexts, means grounding your model in the context it is being implemented in. Resource, infrastructure, community dynamics, and the specific barriers – social, financial, logistical – affect who can access a programme or intervention. Listening to communities is critical.

This is why running our own PEAS schools in Uganda and Zambia matters – we operate in the same rural, low-resource environments as our government partners. That firsthand experience means we’re sharing what we’ve tested, adapted, and seen work, not theoretical recommendations.

Final reflections: Scale what works, not what scales

The session closed with a provocation that is worth sitting with: scale what works, not what scales.

It challenges a default in the sector – the tendency to prioritise models that are easy to replicate, at the expense of approaches that are effective but harder to spread. The implication: be willing to do the slower, harder work of embedding what genuinely works into government systems, rather than optimising for the appearance of scale.

For PEAS, that is the work. Our school improvement programmes in Uganda, Zambia, and Ghana move through 5 phases:

1. Innovate and test in PEAS own schools
2. Collaboratively adapt and codify innovations and tools with government
3. Test, iterate and embed in government schools
4. Transition to local ownership through light touch mentoring and training
5. Government-led delivery at national scale.

It is ever evolving and takes time, trust, and funders who understand that durable change doesn’t follow a short grant cycle. But the alternative – scaling an innovation that looks good on paper but doesn’t survive contact with the system – isn’t really impact at all.

PEAS works with governments in Uganda, Zambia, and Ghana to improve secondary education at national scale. Our own network of 46 schools serves as the evidence base for the approaches we share with Ministries of Education. PEAS school improvement programme is currently being delivered in 470 partner schools, educating around quarter of a million students.