Evidence · 75 Diagnostics

The same patterns.
Everywhere.

We have run the ClusterOS diagnostic across 75 clusters — cyber security ecosystems in Belfast, Tel Aviv, Cheltenham, Singapore, and San Francisco; advanced manufacturing in the Basque Country; the Cambridge tech ecosystem; Orlando's ten innovation clusters; and more. The finding that surprised us most is not what each ecosystem is doing wrong. It is how consistently they do the same things.

75+
Clusters
diagnosed
8+
Countries
& regions
50170
Evidence items
per cluster
9
Canonical
stall types

Coverage

Clusters diagnosed
across geographies

Diagnostics completed across North America, Europe, the Middle East, Southeast Asia, and Australasia. Sectors include cyber security, advanced manufacturing, space technology, medtech, digital media, agtech, photonics, simulation & training, and regional innovation ecosystems.

Geographies represented
Belfast Cambridge Tel Aviv Singapore Basque San Francisco Orlando Full diagnostic Structural resemblance
Belfast Tel Aviv Cheltenham Singapore San Francisco Basque Country Cambridge Orlando + structural resemblances across 15 further ecosystems

Finding 1

Three stalls appear in
almost every ecosystem

Across 75 diagnostics, we tracked which of the nine canonical stalls appeared in each cluster. The distribution is not even. Three stalls dominate — present across virtually every geography, sector, and maturity level we have encountered.

Frequency reflects appearance across completed diagnostics. Evidence confidence varies — stalls with high X-side observability (S2, S7, S8) show higher confidence than those requiring Y-side verification (S4, S1). Numbers indicate approximate proportion of diagnostics in which each stall was identified.

"Narrative × Activity is the most frequently identified stack across all 75 diagnostics. It appears in mature ecosystems and emerging ones. It appears in ecosystems with strong university anchors and in those with corporate anchors. Geography, sector, and maturity level do not predict its presence. Something else does."

ClusterOS Diagnostic Database · 75 completed runs

Finding 2

Three patterns that
cut across everything

These are not sector findings or regional findings. They are system-level patterns — dynamics that appear regardless of what a cluster makes, who funds it, or where it sits.

1
Universal
Coordination is the default response to pressure — in every context
S2 (Coordinating instead of deciding) appears in more diagnostics than any other stall. It appears in early-stage ecosystems trying to build momentum and in mature ecosystems that have been coordinating for decades. When pressure rises — from funders, from government, from competitive threat — the system's first response is to convene, align, and form a working group. In no ecosystem we have diagnosed has the coordination response been absent.
2
Near-universal
The growth-stage capital gap is structural, not accidental
Every ecosystem we have diagnosed has abundant early-stage support. Every ecosystem we have diagnosed has a documented gap at growth stage. Companies that reach Series A or B either relocate or fail to find follow-on capital locally. The system's response — in every case — has been to create more early-stage programmes rather than address the retention constraint. The hole is at the bottom of the funnel. The response is always to widen the top.
3
Cross-sector
Incumbent anchor concentration is the most broadly observed structural pattern
In cyber security, advanced manufacturing, simulation & training, tourism technology, and medtech — in Belfast, Singapore, Orlando, the Basque Country — the same structure appears: large incumbents define innovation pathways, coordination structures form around incumbent needs, new entrants optimise for incumbent partnership, and emergence outside the incumbent orbit remains weakly developed. The incumbents are different. The structure is the same.

Finding 3

The stacks that appear
most often

Individual stalls are informative. Stacks — mutually reinforcing combinations — are where the diagnostic produces its most actionable findings. These are the configurations we encounter most frequently.


Finding 4

Ecosystems on different
continents show identical structures

The most surprising output of comparative diagnostics is structural resemblance — pairs of ecosystems in different countries, different sectors, different political contexts, that show almost identical behavioural configurations. The resemblance is not in what they produce. It is in how the system stabilises.

"Once the stabilisation logic becomes visible, similar patterns can be seen in many other ecosystems — sometimes with different labels, different justifications, and different political dynamics, but with remarkably similar behavioural structure. That recognition is where stewardship begins."

Ecosystem Stewardship — ClusterOS Research Framework

Implication

What the data
means for your ecosystem

Pattern consistency across 75 diagnostics has a practical implication: your ecosystem is more predictable than it looks. The diagnostic does not need to start from scratch. It applies a framework refined across dozens of ecosystems to identify which patterns are operating in yours — and where the specific leverage points are.

Your stack is probably already named
The most common configurations have been seen before, named, and analysed for leverage points. The diagnostic identifies which one is operating — quickly.
The context is yours — the pattern is not unique
The actors, politics, and history are specific to your region. The behavioural structure almost certainly is not. Structural resemblances tell you what has shifted regimes elsewhere.
Single interventions will keep failing
If the pattern data is right, targeting one stall without addressing the stack will produce the same result it produces everywhere: the other stalls compensate.
Small perturbations, not programmes
The leverage hypotheses generated by the diagnostic are not new strategies. They are small withdrawals of protection, tested in ecosystems with comparable configurations.

Your ecosystem is in here.
Let's find it.

The diagnostic identifies your stalls, names your stack, and generates leverage hypotheses calibrated to your specific configuration — informed by every ecosystem we have diagnosed before it.

Request a diagnostic
Testable leverage hypotheses · Informed by 75+ comparable ecosystems
Live example
Orlando Innovation Ecosystem

10 clusters, 582 evidence items, three cross-cluster patterns, two behavioural stacks, and a prioritised leverage hypothesis. A full ClusterOS diagnostic summary — based on public evidence alone.

View summary analysis →