Notes on building an Accelerator for Good across the world: cognitively un-flattening the landscape

One of the core premises of our work in the Curiosity Incubator, and now in building the Accelerator for Good, based on our work in over 40 countries in over 120 projects in the past 10 years, is how to incorporate not just what “worked”, in finding solutions to our most pressing problems.

But rather, incorporating all the incredibly important information, experience,  know-how and capabilities creation of what didn’t work, into the design, content, and work of the Accelerator itself.

One of the reasons why we focus so much on collective intelligence, and co-creation to co-building of solutions, is the dangers present in some of the common misconceptions about how solutions – and people – work, and how they derail the exact work we all need to succeed, to combat the steep challenges we all face.

In this blog, we talk about what we’re learning in real time with the projects and teams in the Accelerator, and we’re looking to show not just tell how some of this is working – what we thought at the outset might work, and how that changed and shifted as the co-creation process began.

Challenge 1: learning from the wrong things

Survivorship bias or survivor bias is a statistical error that results from concentrating on entities that passed a selection process while overlooking those that did not. This can lead to incorrect conclusions because of incomplete data.

Survivorship bias is a form of sampling bias that can lead to overly optimistic beliefs because multiple failures are overlooked, such as when companies that no longer exist are excluded from analyses of financial performance. It can also lead to the false belief that the successes in a group have some special property, rather than just a coincidence, as in correlation “proves” causality.  

The classic example is Abraham Wald’s work on bullet holes in planes coming back from battle in WWII: the planes that returned from battle statistically looked like this, with most of the bullet holes in the wings and tail, so the decision was made to reinforce the wings and tail of planes as they were “obviously” vulnerable to bullet holes.

Beloved of business school curricula, Wald's plane is used to talk about misconceptions about entrepreneurship there, but is an incredibly common cognitive bias that stands in the way of better solutions across the board

Wald said no: reinforce the engine.

Why? Because none of the planes hit there returned (they blew up, and their pilots died). If the army had taken the decision based on the data they had in front of them (that ‘survived’) - the decision the majority wanted to take - they would be reinforcing the exact wrong part of the plane.

The way we commonly approach solutions creation - both what we use to define problems, and the way we approach the process - can be beset by our own form of survivorship bias: case studies, and “learning from examples”.

When we focus on “best practice case studies”, we learn, not how to spot challenges like survivorship bias and reinforcing the exact wrong part of the plane, but only from the examples that “succeeded” ie survived. We learn nothing about how those solutions succeeded, or how they worked, or were built – just a nice story about the end result.

So we learn the exact wrong thing, and it further entrenches people’s overconfidence bias; the belief that we will more easily be able to do build something ourselves, especially something new, than we likely will. Instead we need to be learning how to build different kinds of solutions, with people, from the ground up - and the full complex picture of what didn’t work is more useful in this process, than a finished result of what did.

What do we do instead? 

Learn, not to analyse simplified and flattened landscapes, and come up with pre-packaged “solutions” that will "scale", but how to build, together with people, solution sets, approaches, networks, and tools, that can be used to collectively solve what lies in front of us.

People with the most useful knowledge of how to solve problems are not high-level analysts who have never been to the places they are prescribing one-size-fits-all solutions for, they’re people closest to the challenges and their impacts themselves, together with the people with the most experience in trying to deliver previous iterations of solutions that didn’t work, or partially might have worked.

What we can do is come up with frameworks of what might work – based on rigorous, detailed knowledge of the problem sets, and the complexity of what the terrain might be. Then those frameworks, which are built to be continually tested by and with people, can flourish, adapt, and work, in each of the specific contexts and circumstances of Places where people are.

In public policy terms, that would be people experiencing the sharp end of the challenges, together with front-line workers who work with them from any sector, who have the experience of having tried lots of previously prescribed top-down “solutions”, and learnt the hard and long way, what works and what doesn’t. It’s the road miles that matter, in terms of useful information and skills – not anyone’s ability to write a report or form a clever prompt.

A different kind of learning, a different kind of work

Co-creating solutions that really work for complex challenges requires a different kind of work than we are often putting into them, though this is uneven in different levels of government and society, and different parts of the world [spoiler: other countries especially in the global south are much more skilled and experienced at this than we are in the eg the UK, from my experience of working in each, for example].

Fundamentally, solutions that work do so because of multiple kinds of collective intelligence, co-design and many different kinds of people being part of not just the implementation of one simple top-down solution, but in the cycle from problem definition through iterations of solutions prototypes, onwards. And the solutions co-creation process doesn’t “end”, per se – it’s a continuous process, as continuous as the grappling with the problem itself.

That’s why solutions builders – and landscape navigators – are critical other skillsets and capabilities to develop, as we continuously assess the solutions and how to make them work, together.

Accelerating for Good – our newly-launched, organisation/team-based 6-month long co-creation and collective intelligence approach to solving our most pressing problems: we’ll blog as we go, and further background here.

Onwards!