Scale Is Not a New Question

It must be Skoll Forum week soon because my timeline is full of debates about scale – whether to scale, what to scale, how to scale, who scale leaves behind, etc. 🙂

Which isn’t to say these aren’t serious questions. But are they really new?

Because this debate has happened before. The Basic Needs movement in the 1970s challenged the growth-led paradigm. The participatory development turn in the 1980s challenged the top-down delivery model. Each time, the challenge was partially absorbed and partially neutralised. The delivery model adapted its language without changing its logic, reducing participation to a checkbox and local leadership to a procurement category.

Something similar is at risk of happening again. The current debate is still, at its core, a debate about delivery. Whether to scale, what to scale, and how to scale — these are all questions about the intervention.

What almost nobody is asking is what the system needs to be capable of receiving what we’re trying to deliver.

The binding constraint isn’t the model. It’s the relational substrate — the trust, the reciprocity, the shared meaning that accumulates between people and institutions over time. You can’t deposit exponential impact into a system whose connective tissue has been decayed by the very logic you’re now trying to scale through it.

In agriculture, nitrogen fertiliser produces dramatic short-term growth. Without phosphorus, the soil can’t absorb it, and repeated application without attention to the soil’s underlying biology eventually degrades the capacity you were trying to enhance. The intervention becomes the cause of the limitation it was trying to overcome.

The relational substrate doesn’t have a metrics infrastructure, it doesn’t have a funding category, and it doesn’t have institutional-scale champions willing to say bluntly that this is the binding constraint.

Not how do we scale better? What does the system need to be capable of receiving what we want to give it?

The first question has become more sophisticated over the past fifty years, while the second one has barely been asked.

Originally written for LinkedIn on 18 April 2026. View original →