Why Two Providers Disagree About the Same Submarket
Submarket boundaries are editorial decisions. Comparing a statistic across providers usually means comparing two different geographies.
Data & analytics · 488 words · updated 2026-08-26
Market and submarket geographies are not standardised. Each data provider draws its own, and those choices propagate into every statistic keyed to them. This is the most common reason two credible sources report different vacancy rates for what appears to be the same place.
Five choices that create the divergence
- Boundary source. Some providers build submarkets from municipal boundaries, some from census geography, some from broker convention, some from drive-time or transport corridors. None is wrong; none is comparable to another.
- Inventory threshold. Whether a building is tracked at all commonly depends on a minimum size — often somewhere between 5,000 and 25,000 square feet — and on whether owner-occupied, government-owned, medical, and single-tenant build-to-suit product is included.
- Class assignment. "Class A" is a relative judgement about a building's position within its own submarket, not an absolute standard. It drifts as new supply arrives, so a Class A series can shift without any building changing.
- Space status treatment. Whether sublease space counts in availability, whether space under a signed but uncommenced lease counts as occupied, and whether under-renovation space stays in inventory.
- Measurement standard. Which BOMA revision the rentable areas follow, and whether they were standardised at all.
What to do about it
- Pick one source per question and stay with it. The trend within a consistent series is reliable even where the level is not comparable across series.
- Compare levels only against the same provider's own history. Never mix providers inside one time series.
- Reconcile at the building level when the stakes justify it. If two sources disagree materially, the reconciliation is a list of specific buildings each includes and the other does not. That list is usually short and immediately explains the gap.
- Read the methodology note. Inventory threshold, class definition and sublease treatment are usually disclosed, and those three explain most divergences.
- State the definition alongside the number in anything you publish. A vacancy rate without its inventory definition is not a reproducible figure.
Where it matters most
Divergence is largest where inventory is small and one large building dominates. In a submarket with two million square feet, a single 400,000 square foot building entering or leaving the tracked set moves vacancy by several percentage points. Treat small-submarket statistics as indicative and go to the building list.
It matters least for direction. Providers with different geographies and thresholds nearly always agree about whether a market is tightening or loosening, because the underlying leasing activity is the same. Use levels for context and changes for decisions.
Custom geographies
For most real decisions the right geography is neither the provider's market nor its submarket, but the one the decision lives in — a drive-time isochrone around a site, a corridor, a radius, a set of specific parcels. Building statistics on a custom geography from building-level data avoids the whole problem, which is why building-level access matters more than the polished submarket aggregate sitting on top of it.