In late 2024, Scottish midfielder Scott McTominay moved from Manchester United to Napoli. He was an elite athlete, at the peak of physical performance, who had spent all his life in the UK. Yet, within weeks of arriving in Southern Italy, he went viral for a comment about a tomato.

“I never ate them at home,” he told The Athletic. “They’re just red water. Here, they actually taste like tomatoes… I eat them as a snack now. It’s incredible.”

The tomatoes McTominay experienced in the UK weren’t mediocre by accident; they were bad by design. To make a tomato a global commodity, the industry had to solve for friction:

The Problem: Real tomatoes are fragile, oddly shaped, and have a shelf life of days.

The Optimisation: Breeders developed tough varieties that can be picked green, gassed with ethylene to turn them red on command, and survive 2,000 miles in a vibrating truck.

The result is a miracle of logistics. You can buy a perfectly spherical, bright red tomato in Manchester in December for pence. But as McTominay discovered, when you solve for optimisation, you often destroy the content. By removing the friction of fragility and spoilage, we also remove the flavour. We replace a complex, acidic sweetness with “red water.”

This isn’t just true about tomatoes, it’s a template for how we can ruin place based innovations.

The Harlem Children’s Zone (HCZ) is a cautionary tale of what happens when we try to scale an innovative practice.

HCZ started as a hyper-local, street by street effort in Harlem to wrap children in a ‘cradle-to-career’ support system. Its success wasn’t just in the services provided (after-school programmes, health screenings, college preparation) but in the relational infrastructure that Geoffrey Canada and his team spent years building. It was a dense, trust-based ecosystem nurtured by the culture and history of a specific neighbourhood.

When the federal government attempted to scale this nationally via the “Promise Neighbourhoods” initiative, they committed the cardinal sin of a lift and drop. ” They took the HCZ checklist—the services, the funding streams —and asked other cities to replicate them.

Other cities implemented the form of the programme, but missed the soul. They tried to manufacture street level results through top-down policy, failing to realise that you cannot mandate the human connection that makes a place-based initiative truly work.

We are at a critical point in developing the next stage of place-based working at BFL. The pilots that we introduced have been validated and we must move to the next stage: preparing to mobilise circa 100 place based teams.

So do you scale them or spread them?

Spreading (or diffusion) is a horizontal, social process driven by interpersonal influence, peer-to-peer networking, and social replication. Spreading embraces local adaptation allowing people the opportunity to test and adapt the change to fit their unique local environment. Spreading is highly effective for practice improvements, behavioural shifts, and initiatives requiring deep cultural alignment.

Scaling (or managed scale-up), on the other hand, is a deliberate, vertical, and coordinated effort to expand the reach of an intervention. Scaling is typically a top-down, mechanical process that requires tackling infrastructural challenges, such as building necessary technological, procedural or changing management layers. Unlike the flexible nature of spreading, scaling generally demands high fidelity and strict adherence to the original model to guarantee consistency. It is best suited for structural innovations, tech systems, or highly regulated practices (say, a new medicine.)

When organisations try to scale a complex, relational practice using purely managed, top-down methods the results are often disastrous. This approach overwhelms organisational capacity, misaligns resources, and creates internal conflict. More importantly, it eliminates the opportunity for localised trial-and-error, exposing minor hiccups to the entire system and instantly eroding the fragile community trust that place-based models rely on.

This is the risk of our expansion. That our place roll-out is co-opted by bureaucracy, diluting a vibrant locally-driven initiative into a standardised, low-fidelity compliance exercise.

In the world of scaling innovation, we are obsessed with the delivery mechanism. We want our ideas, our cultures, and our programmes to survive the journey from the head office to the front line without deviation. To make that happen, we often design them for durability: we strip away the weird, fragile, human elements—the local trust, the specific expertise, the messy relational work—because those things are hard to pack and even harder to ship.

We end up with a version of the innovation that is perfectly uniform, easy to roll out, and completely, utterly bland. We have optimised for the logistics of the project, but we have starved the experience of it.

That’s our big strategic innovation challenge: avoiding the watery tomato.