Every figure on SuburbIQ traces back to a public dataset. Here is exactly what we use, what we do to it, and where it can mislead you.
Rent figures come from the rental bond dataset published by the Ministry of Business, Innovation and Employment through Tenancy Services. Landlords are legally required to lodge tenancy bonds, so the dataset records rents tenants actually agreed to pay — unlike listing sites, which record what a landlord hoped to get.
We use the detailed quarterly file, which reports each statistical area by dwelling type and bedroom count from 2020 onward. Geographic names and boundaries come from Stats NZ: the Statistical Area 2 (2018) classification, joined to territorial authorities through the Meshblock Higher Geographies concordance.
Official statistics divide Auckland into statistical areas, not suburbs. Larger suburbs are split across several — Mount Wellington alone is seven. Nobody searches for “Mount Wellington North West”, so we group these areas back into the suburb people actually use.
Where a suburb spans several areas we combine their medians using a bond-count-weighted geometric mean. Weighting by bond count stops a tiny area with three tenancies swinging the result; a geometric rather than arithmetic mean suits rent data, which is skewed to the right by a small number of very expensive properties. Where a suburb is a single statistical area — about half of them — the figure we publish is the published median, unchanged.
MBIE applies fixed random rounding to base three and withholds any cell built on fewer than five bonds, protecting the privacy of individual tenants. That is why some suburbs show a dash for four- and five-bedroom properties, and why very small suburbs do not appear at all. We publish a suburb page only when it has at least 15 new bonds in the latest quarter.
Every page shows the bond count behind its median. Treat a median built on 20 bonds as indicative, and one built on 400 as solid.
These figures describe new tenancies only. Rents on existing tenancies, typically lower because they have not been reset to market, are not represented. The quarterly figure is closer to “what it costs to move in today” than “what the average tenant pays”.
Property mix matters more than people expect. A suburb with many single rooms or student flats shows a much lower median than its houses alone, which is why every suburb page breaks the figure down by property type and flags the cases where the mix distorts the headline.
Our Auckland-wide pooled median sits within about 2% of the figure MBIE publishes directly for Auckland. The gap exists because a small share of bonds cannot be assigned to any statistical area, and because combining area-level medians is not mathematically identical to taking the median of every bond at once. Where you need the single official Auckland number, use MBIE's.
Suburb boundaries follow Stats NZ definitions, which occasionally differ from how locals or real-estate marketing draw them. “Nearby suburbs” are ranked by straight-line distance between area centres, so a suburb across a harbour can appear closer than it feels to drive. We publish no property-level data, no addresses, and no council valuations.
MBIE releases the detailed quarterly file each quarter. SuburbIQ is rebuilt when the new file lands, refreshing every page. The period covered is stated on every page.
Rental bond data © Ministry of Business, Innovation and Employment, reused under Creative Commons Attribution 3.0 New Zealand. Geographic classifications © Stats NZ, reused under Creative Commons Attribution 4.0. SuburbIQ is an independent project, not affiliated with or endorsed by either agency.