Ecosystems don't photograph well
The problem with funding the things that work
A few years ago, Hilda Bugingo’s sister gave birth to premature twins in Uganda. Two days after the birth, the hospital ran out of oxygen and both babies died. Hilda tells this confronting story because it’s why she does what she does now: working as an engineer for FREO2, building the systems that keep oxygen flowing in hospitals.
Post Covid, a huge amount of money flowed into Sub-Saharan Africa to try and solve the oxygen crisis. In Nigeria alone, more than 100 oxygen production plants were built. So, how is it that people are still dying for lack of what so many of us take for granted?
Simply put, just because the equipment is there, doesn’t mean patients are able to access oxygen. What’s missing is the connective tissue: the network of local businesses which would maintain the equipment, manage delivery and keep supply running when the power goes out. And it’s missing because it doesn’t get funded.
Over the past five years Brink has been working on the Oxygen CoLab with our partners DT Global, UNICEF and our funders at the FCDO, proving that backing these businesses with the right capital and support, results in the delivery of 100% uninterrupted oxygen to patients at a cost up to 70% lower than existing methods. This week we launched a campaign asking funders to shift how and where they invest.
Why hasn’t this shift happened already? There are a lot of complex factors at play - entrenched paradigms, politics, neo-colonialism, etc - but as in all things, when mindsets shift, everything else becomes possible. So for this newsletter we wanted to zoom in on two cognitive patterns that we think are keeping funders stuck in the same groove, and why cracking them might be key to solving problems far beyond oxygen provision.
The attribution problem
Attribution is easy when impact is visible and concentrated. If the donor funds the plant, then the plant produces oxygen, and patients live. That’s a nice, neat causal chain. But for oxygen to actually reach a patient there needs to be procurement reform, performance-based contracts, maintenance protocols, logistics networks, training programmes and blended finance that helps operators scale before commercial capital arrives.
In systems work there’s no lone hero, just an ecosystem. And ecosystems don’t photograph well at ribbon-cutting ceremonies.
Which part should get credit when a patient survives? All of them, which in essence means none of them. In systems work there’s no lone hero, just an ecosystem. And ecosystems don’t photograph well at ribbon-cutting ceremonies. That’s the attribution problem.
But tangled up with it is an accountability problem, and arguably that’s the harder one to solve. A survey of around 100 organisations doing systems-change work found that most measure progress through shifts in policy, changes in relationships across networks, or movement in the narratives that shape how a problem is understood. These are the indicators that actually tell you whether entrenched systems are starting to move. Only 17% rely on traditional metrics like ‘number of people served’, because for this kind of work, that number tells you almost nothing.
Crucially, the same survey found these organisations are consistently underfunded, largely because what they measure doesn’t fit the accountability structures funders have built.
The narrative problem
As we’ve written before, stories trigger the brain’s chemistry in ways that data alone can’t. They release the cocktail of oxytocin, dopamine and cortisol that makes information stick and prompts action. That’s why it’s so frustrating when certain projects turn out to be stubbornly resistant to narrative.
Stories need a protagonist and a moment of transformation; a distinct ‘before and after’ that you can point to. But it’s hard to fit complex maintenance systems into that equation.
Consider FREO2’s work in Uganda: during a three-month trial, one hospital experienced 215 power outages and oxygen never ran out. That’s a remarkable result… Which is also teeth-grindingly tricky to turn into a catchy story. How do you turn the continuous, invisible work of a system functioning as intended into a compelling narrative?
In the past, when we’ve tried to communicate what the Oxygen CoLab does, it’s been too easy to reach for technical jargon or lean on tales of individual heroics. But both of those miss the actual intervention, which is a whole stack of interlinking actions that together create something none of them could achieve alone.
Research on systemic storytelling suggests you need “parallel story arcs” that focus “not on one but several heroes and bring a broader perspective”. Parallel story arcs are what we created to tell the story of the Oxygen CoLab (although that’s not what we called them at the time!).
We know that global health funders need independent academic evidence before they move, so we pursued that route. But we also know that process takes time and can’t capture the kind of insights you get when a model meets the real world. So we ran both tracks at once, building the evidence base in real time and sharing what we were learning as we went. And rather than trying to tell the system’s story ourselves, we handed the microphone to the people whose lives it had changed.
Going Beyond Oxygen
Are these issues of mindset the reason why 91% of people in Sub-Saharan Africa who need oxygen for acute medical or surgical care still aren’t getting it? Of course not. But funders need to be able to see where they fit in a story, and how investment in systems changes outcomes. When that motivation becomes obvious, other barriers become surmountable.
Oxygen is where we’re testing this approach right now, but the underlying argument extends well beyond it.
Most of the problems that have resisted decades of investment share the same structural features: impact distributed across networks rather than concentrated in visible outputs, and change that accumulates too slowly to fit inside annual reporting cycles. Think of the billions spent on solar panels while the grid infrastructure needed to make renewable energy reliable remains chronically underfunded. Or the rush to put AI tools into classrooms across low-income countries, while the teacher training, data systems and local capacity that determine whether those tools actually work for the children in front of them gets a fraction of the attention.
Getting unstuck will require funders to value and measure that kind of work differently, but it will also require all of us to find new ways of seeing complex systems clearly enough to be able to tell honest stories about them, and to build the narrative and attribution frameworks that make invisible infrastructure visible, fundable, and worth celebrating.
This month’s mystery links
Your reward for reading down this far…
There are currently 1,701 people receiving this newsletter. NCC-1701 is the registry number of the USS Enterprise 🤓. Its mission, famously, was to boldly go somewhere, not knowing what it would find there. But Starfleet funded the voyage anyway!
If you’d like to help us keep this newsletter boldly going (or going boldly, even), then click the button to share with a friend, colleague or local Vulcan: