The answer first. Site selection is a strategy decision wearing a real estate costume. Choosing the right site for a factory was worth CAD 39 million over ten years — not because of cheaper land, but because of who was next door. The winning site sat within 5 kilometres of six aerospace primes, in a cluster with a 4.428 advanced-manufacturing location quotient and a 1.59:1 tech-to-hardware density ratio. The moment you hand site selection to the real estate team to run as a cost-minimization exercise, you have already lost the value that location was supposed to create.
The finding: access beats cost, and it is measurable
The standard playbook ranks sites on cost inputs — land, labour, tax — and treats location as a line item to minimize. That is backwards for any business where access matters more than cost: defence, aerospace, biotech, semiconductors, anything regulated or ecosystem-dependent.
The winning site’s advantage was not cheaper rent. It was the cluster:
| Signal | The number |
|---|---|
| Aerospace primes within 5 km | 6 |
| Advanced-manufacturing location quotient | 4.428 (~4x North-American density) |
| Tech-to-hardware density ratio | 1.59:1 |
| A&D organizations in the cluster | 85+ |
| Tech workers / STEM students | 29,000+ / 42,000+ |
That proximity did three things, none of which show up in a rent-per-square-foot comparison: it compressed certification (the regulators were already there), cut logistics cost (suppliers were minutes away), and de-risked hiring (the skilled workforce was already concentrated).
The framework: the neighbour test
The neighbour test is a single question asked before any cost comparison: what do I need to be near — and is that worth more than the rent I would save being somewhere else?
Two people look at the same map and see different problems:
| Cost-first view | Access-first view |
|---|---|
| Cheapest land | Closest to the primes and regulators |
| Lowest labour cost | Deepest skilled-labour pool |
| Lowest tax | Fastest certification path |
The CAD 39 million — a CAD 5.7M/year fully-loaded OpEx advantage compounded over ten years at 8% — is the quantified difference between those two views. It is the price of asking the wrong question.
The trap: handing the map to the wrong person
The mistake is structural, not analytical. The site-selection playbook hands the map to the real estate team, which optimizes cost inputs. For a regulated or ecosystem-dependent business, that optimization destroys the value location was supposed to create.
The tell: if your site-selection model has no line for “who is next door,” it is measuring the wrong thing.
The decision agenda: what to do differently
- Run the neighbour test first. Name the neighbour that matters — the prime, the regulator, the talent pool — before you compare rent.
- Model access, not just cost. Certification timelines, logistics, and hiring all show up in the P&L; none of them show up in the lease.
- Treat location as a strategy input, not a cost line. For access-dependent businesses, the cheapest site is often the most expensive mistake.
What could break this read
- The CAD 39M is modeled. The 8% WACC and 10-year horizon are assumptions. The direction — co-location beats cheap land — is robust; the exact number is a planning figure, not a guarantee.
- You can buy access without moving. A partnership or a satellite office near the primes can capture some of the certification and hiring benefits without relocating the whole factory. Co-location is the strongest form of access, not the only one.
The checklist
The rule: name the neighbour that matters before you compare rent. Site selection is a strategy decision, not a real estate one.
Sources
- 15Rock site-selection analysis (2026) — CAD 5.7M/year fully-loaded OpEx advantage, ~CAD 39M cumulative 10-year savings (8% WACC); six-prime co-location within 5 km.
- Aerospace cluster sector scan (2026) — 4.428 advanced-manufacturing location quotient, 1.59:1 tech-to-hardware density ratio, 85+ A&D organizations, 29,000+ tech workers, 42,000+ STEM students.
- Canada’s Defence Industrial Strategy (Feb 2026) — raise domestic procurement share to 70%, grow defence industry revenues 240%.
Figures are modeled estimates or reported from the site analysis. Client and counterparty names are withheld.