Schools, childcare, neighbourhoods - all in one tool.

Australian Bureau of Statistics (ABS) Regional Population - Estimated Resident Population (ERP), year ending June 2025, at SA2 (Statistical Area Level 2) level. Fields sourced from this release: population, 1-year growth %, growth components (natural increase, net internal migration, net overseas migration), population density, and SA3/SA4 hierarchy. SA2 boundaries don't always equal suburb boundaries and the relevant SA2 is named on each profile.

The forward-looking population figures are the state governments' own published projections, republished as released: Victoria in Future 2023 (Victorian Department of Transport and Planning, released September 2023, to 2036) and Travel Zone Projections 2024 (Transport for NSW, to 2066 - we show to 2041), both free under CC BY 4.0. We publish no forecast of our own. Only the years each publisher lists are plotted: we never interpolate between their horizons, extend past the last one, blend the two states, or turn them into a growth rate, index or score. Both states publish at local government area (council) level, which is larger than a suburb - no state publishes a suburb-level projection - so we resolve each area's council from the ABS 2021 meshblock allocation, weighted by Census population, and label the figure as the council area throughout. Figures are rounded to the nearest 10. A projection is an assumption about the future, not a measurement of it, and it is revised at each release; it sits on a different series to the ABS resident-population and Census counts elsewhere on the page and is not directly comparable with them. Other states are shown nothing rather than a substituted figure.

The "what's being built here" figures are ABS building approvals - the number of dwelling units councils approved - republished per SA2 for each year ended 30 June, from ABS Data by Region (cat. 1410.0), which carries them from ABS Building Approvals, Australia (cat. 8731.0). Free under CC BY 4.0. An approval is not a completion: some approved dwellings are never built and the rest take years, so read this as what is planned around a suburb, not as homes standing in it. It is also not a forecast - we project nothing forward. The only arithmetic we apply is adding the ABS's own yearly counts together, and the page says which years it added; we never divide approvals by land area or population, and we compute no "development pressure" score, rate or index. The total covers public and private dwellings; the separate-houses and townhouses-and-apartments figures are the private-sector split the ABS publishes, so the two parts do not add up to the total wherever public housing was approved, and we never present them as if they did. Where the ABS publishes no figure for a year, we draw no line rather than closing the gap. The state comparison line is the median area in the same state with a complete series - a typical area for context, never a target. Why not council development applications: the only Australia-wide DA feed, PlanningAlerts, licenses its free tier for strictly non-commercial use and prices commercial access at A$3,850 per month, and individual council portals publish on hundreds of different schemas and licences - so a DA feed would break the free-data rule, while ABS approvals count the same thing everywhere.

The suburb outlines drawn on our maps are the ABS Suburbs and Localities (SAL) boundaries from the Australian Statistical Geography Standard Edition 3 (2021) - the gazetted suburbs and localities the states and territories actually name, released free under CC BY 4.0. They are a different geography to the SA2 statistical areas our figures are published at: one SA2 named "Cobblebank - Strathtulloh" covers two separate gazetted suburbs, and the map now draws both. Where a figure is published per SA2 - including the Family-Fit score and SEIFA - every suburb inside that SA2 carries the same value, and we say so next to it rather than implying the number was measured for the individual suburb. The other metrics you can shade the map by - median age, children under 15, born overseas, average school ICSEA and median house price - are read from each gazetted suburb's own Census, ACARA and Valuer-General records, so neighbouring suburbs genuinely differ. The map's legend names the active metric and states which of the two it is. The same mismatch shapes the search box: an SA2 is named after only some of the places inside it, so searching for a gazetted suburb our page titles never carry - Harkness, Umina Beach, Beecroft, and about 4,100 others of 200 residents or more - now resolves to the profile whose area contains it, marked "Part of our … profile" so it is clear the page describes the wider area rather than that suburb on its own.

ABS Census 2021 General Community Profile (median age, median household and personal income, overseas-born %, language background, Indigenous %, children aged 0–14, and the full age structure in the ABS's own age groups - combined only where whole groups sit side by side, never split) and ABS SEIFA 2021 (IRSD, IRSAD, IER and IEO deciles plus the IRSAD score), both at SA2 level. Census figures reflect the 2021 collection night; the next refresh follows the Census 2026 release.

Australian Curriculum, Assessment and Reporting Authority (ACARA) School Profile and School Location 2025 releases: 9,755 schools with sector, type, year range, enrolment, ICSEA and location. ICSEA reflects the academic and socioeconomic context of a school's student intake - it is published by ACARA, not computed by us. We do not display NAPLAN results: per-school NAPLAN data sits behind ACARA's data-access application process, and we won't show it without a clean licence. School catchment boundaries come from the Victorian Department of Education (School Zones 2027) and the NSW Department of Education (School Intake Zones 2026); other states' catchments will be added as their departments publish open data.

Victorian Curriculum and Assessment Authority (VCAA) Senior Secondary Completion and Achievement Information (SSCAI), 2025 release, accurate as at 6 December 2025. We show median VCE study score as one published measure of senior-secondary outcomes - never a ranking or composite score, and never for primary schools. Small cohorts are suppressed by VCAA and shown as "not published" here, exactly as they appear in the source. Used with permission. © Victorian Curriculum and Assessment Authority; reproduced with permission. For current versions and related content visit http://www.vcaa.vic.edu.au

Australian Children's Education & Care Quality Authority (ACECQA) national registers: 18,000+ education and care services with National Quality Standard (NQS) ratings, service type, approved places and address. The register is refreshed daily by ACECQA; our extract is updated periodically and each page shows its data date.

Department of Education Child Care Subsidy (CCS) quarterly data report, December quarter 2025: mean hourly fee for Centre Based Day Care and Outside School Hours Care, already aggregated by the Department. Published at SA3 level - a wider area than the SA2 each suburb page describes - so the same figure is shown across every suburb in that SA3. No SA3-level breakdown exists for Family Day Care, so it isn't shown. © Commonwealth of Australia, CC BY 4.0.

Victoria: Valuer-General Victoria median house prices (Q4 2025) and DFFH moving-annual rents (Sep 2025 - the latest quarter DFFH has published). NSW: Department of Communities and Justice sales (Mar 2026) and rent tables (Jun 2026). All open government releases. Other states are not yet covered - pages show this explicitly rather than substituting modelled values. We do not use or redistribute proprietary CoreLogic, Domain or REA data.

NSW: Bureau of Crime Statistics and Research (BOCSAR) recorded criminal incidents by suburb (calendar year 2025, from the 2026 Q1 release - the current year is still partial and is deliberately not published as an annual total). Victoria: Crime Statistics Agency criminal incidents by suburb/town (year ending March 2026). We republish the official incident counts with their categories - counts are not population-adjusted, we compute no safety score, and NSW and Victorian figures are not comparable to each other because police recording practices differ.

OpenStreetMap (© OpenStreetMap contributors, ODbL) point counts within each SA2 polygon: general practitioners and clinics, supermarkets, parks, train stations, places of worship, libraries, community centres and playgrounds. OSM coverage varies by area and is contributor-maintained - useful for relative comparison, not a guarantee that every local service is listed. Baby-change table and baby-care-room counts come from a separate source, the National Public Toilet Map (Australian Government Department of Health, Disability and Ageing, CC BY 3.0 AU) - a public-toilet register, not an exhaustive family-facilities list.

Median scheduled weekday AM-peak train travel time to the state CBD terminus from the nearest station, computed from each state's static GTFS feed (Transport Victoria, TransLink, Transperth). Victoria includes both Metro (to Flinders Street) and V/Line regional trains (to Southern Cross) - Brisbane uses Central station, Perth uses Perth Stn/Underground. Train only - bus and tram are not in scope. NSW, SA and ACT GTFS bundles require a free developer-account signup at their respective transport portals; we'll add those states once registered.

When you type a 4-digit postcode in the Safety search, we resolve it to the suburb name(s) the crime data uses. Source: Queensland Government "Locality Lookup" (CC-BY 4.0), itself a re-publication of the community-maintained matthewproctor/australian_postcodes dataset (April 2025 snapshot). Covers all active Australian postcodes; we currently surface NSW and Victoria only because that is the coverage of our crime dataset.

Victoria: whether an SA2 sits within an official Victorian planning-scheme overlay - the Bushfire Management Overlay (BMO), the flood overlays (Land Subject to Inundation, Floodway, Special Building), the Heritage Overlay (HO), the Airport Environs overlays (AEO and the Melbourne Airport MAEO) and the Environmental Audit Overlay (EAO) - and what share of the SA2 area each covers, plus, from the same source, whether it intersects the state's Coastal Acid Sulfate Soils mapping (the Prospective class only; the layer's other two classes are open water and made ground, which describe neither). Source: Vicmap Planning, © The State of Victoria, Department of Energy, Environment and Climate Action, CC-BY 4.0. Alongside the BMO we also publish the designated Bushfire Prone Area (source: Designated Bushfire Prone Area, © State of Victoria, Department of Transport and Planning, CC-BY 4.0). These are two different instruments, not two versions of one: the BMO is a narrow planning control, while the Bushfire Prone Area is the far broader designation that makes bushfire construction standards apply to a new home. Almost all BMO land sits inside the Bushfire Prone Area, but the reverse is not remotely true - 370 Victorian areas are inside the designation against 238 inside the overlay - so reporting only the overlay let an area that is in fact designated read as "not in the overlay". The designation covers 76 councils; the ones it misses are inner-Melbourne councils with no bushfire-prone vegetation, so "not designated" there is a real answer. Overlays are matched on the planning scheme's parent code, so every numbered schedule of an overlay counts (LSIO1, BMO2, HO148 and so on) - an earlier version of this pipeline matched only the bare code and undercounted flood overlays by roughly 60%, which is fixed. New South Wales: whether an SA2 intersects designated Bush Fire Prone Land (source: NSW RFS Bush Fire Prone Land, © State Government of NSW and NSW Rural Fire Service, CC-BY 4.0) and whether it intersects land subject to heritage conservation under the relevant local environmental plan (source: Environmental Planning Instrument – Heritage, © State of New South Wales, Department of Planning, Housing and Infrastructure, CC-BY 4.0; the layer covers 127 of the 128 NSW councils, so an area with none mapped is a real answer) and whether it intersects a mapped aircraft-noise contour (source: Environmental Planning Instrument – Airport Noise, same department and licence) and whether it intersects land classified for acid sulfate soils (source: Environmental Planning Instrument – Acid Sulfate Soils, same department and licence; written into the plans of 52 coastal and estuarine councils). The NSW layers are queried as a yes/no intersection, so they carry no coverage share. The aircraft-noise layer is the one place where an absence is not a real answer: only four councils and the Western Sydney Airport precinct have written a contour into their planning instrument, and Sydney Airport's contours are published federally in its master plan rather than in the state layer - so we label areas with none mapped "Not mapped" rather than "Not in the overlay", and never as quiet. Note that BFPL is mapped broadly across NSW - it designates land carrying or near bushfire-prone vegetation, so most non-inner-urban SA2s intersect it; read its presence as context, not a risk level. In both states we republish the official overlay extent only; we do not compute a bushfire or flood risk score, and this is not insurance or financial advice. Overlay coverage at the SA2 level does not tell you whether a specific address is affected - always check the current planning scheme and your insurer.

The Risks list on each suburb page gathers nine published records in one place: flood overlay, bushfire overlay, heritage overlay, major roads, aircraft-noise contours, land-contamination audit overlay, acid sulfate soils, recorded crime and social housing. Nothing on it is rated, scored, graded or ranked - each badge states what the official record says ("In an official overlay", "Not in the overlay", "Not published"), never how bad it is, and there is no traffic-light colour anywhere on it. Because a missing line would read as "nothing to worry about", every line always renders, including the ones we cannot answer. Social housing is the ABS Census 2021 tenure-and-landlord-type table (G37): the share of occupied private dwellings with a stated tenure that are rented from a state or territory housing authority, and separately from a community housing provider. The two are published separately by the ABS and stay separate here, and renters whose landlord type was not stated sit in the denominator but in neither category - so both figures are floors, not a complete count of social housing. The state comparison is the median reporting area in the state (the census export carries the percentage but not the dwelling base, so there is no honest weight to apply, and a handful of areas above 50% would drag a mean somewhere no typical area sits); it is suppressed below 10 reporting areas. Heritage is an official planning overlay rather than a hazard, and is worded as neither good nor bad: it protects a streetscape and restricts what an owner may change. It applies property by property, so an area inside an overlay is mostly made up of homes that are not. Aircraft noise is a planning control too, not a measurement: the overlays mark land where forecast noise contours around an airport add building and subdivision requirements, and we publish no decibel figure, noise estimate or quiet/loud grade. Land contamination is the Victorian Environmental Audit Overlay: land whose known past use - a service station, dry cleaner, factory or landfill - means an environmental audit must be completed before it may be used for a sensitive purpose such as a home, a childcare centre or a school. It is a planning control that flags land to be checked, not a finding that the land is contaminated, and an audit may find it clean; we publish no contamination measurement of any kind. It is part of the statewide planning scheme and 76 of Victoria's 79 councils apply it, so "not in the overlay" is a real answer there. New South Wales records contaminated land in the EPA's searchable register of notified sites rather than in a planning-scheme map layer, so we show no overlay for NSW and label it "Not mapped" - that is a gap in what we publish, not a statement about the land. Acid sulfate soils come from two different instruments - Victoria's state soil-prospectivity mapping (where the soils are likely to occur; it is soil mapping, not a planning overlay) and the acid sulfate classification written into New South Wales coastal local environmental plans, where work disturbing the soil below a set depth needs consent. Both cover coastal and estuarine land only, because that is where these soils form, so "none mapped" inland means the area sits outside the mapped zone rather than that its soil was tested. We publish no soil measurement, class average or grade. Bushfire is reported from two Victorian instruments rather than one - the Bushfire Management Overlay (a planning control) and the wider designated Bushfire Prone Area (the building designation behind bushfire construction standards) - because an area outside the overlay can still be inside the designation, and showing only the overlay made that read as nothing to check. New South Wales' Bush Fire Prone Land is the equivalent of the designation, not of the overlay, so the two states' figures are not comparable and we never place them side by side. Flood, traffic and crime sources are as described in their own cards. We do not yet cover EPA notices, mine subsidence, landslide or coastal erosion - a line we don't show is a line we haven't checked, not a clear result.

The traffic line names the major roads that run inside a suburb's boundary, read from OpenStreetMap road classifications (© OpenStreetMap contributors, ODbL) - the same free, contributor-maintained map our amenity counts come from, and the only source that classifies roads the same way in all eight states. A road counts for an area if any part of it runs inside the boundary. We group OSM's motorway and trunk classes as "freeway or highway" and primary and secondary as "main road", with on- and off-ramps folded into the road they serve. Roads are counted by distinct name, never by mapped segment: OpenStreetMap splits a road into a new segment at every junction, so a segment count would measure how finely an area was mapped rather than how much road is there, and a suburb with more intersections would look busier on identical roads. Unnamed roads are held as presence only - we never invent a name for one. This is not a noise measurement. We compute no decibel estimate, no noise or traffic score, no road length and no roads-per-square-kilometre figure, and there is no quiet/busy grade anywhere: what a specific home hears depends on its distance from the road, traffic volume, speed, barriers and terrain, and no free dataset publishes those. Because OpenStreetMap is contributor-maintained, a road may be missing or classified differently to how a local would describe it, and traffic on a road just outside the boundary still carries. Every suburb we searched shows a line - including "No major road mapped", which is a result, not a gap.

The Lifestyle section on each suburb page gathers the amenity counts, the nearest emergency department and the train travel time into one place, and shows each count next to the middle area in the same state so a bare number means something. That comparison is the median count across the areas in that state for which we hold that category - median rather than mean because a city-centre area with dozens of tagged parks would drag an average somewhere no ordinary suburb sits; it is suppressed below 10 reporting areas, and where there is no comparison we show none rather than a zero. Counts stay raw OpenStreetMap counts inside the area's boundary: we do not normalise them by area or population, and we compute no lifestyle, walkability or amenity score. Areas differ enormously in size, and OSM is contributor-maintained, so a zero can mean "not mapped" rather than "not there" and a shop just over the boundary will not appear even if it is the one you would use. Distances to the nearest station and hospital are straight-line from the centre of the area - never a walk or drive time, because we hold no routing data. Every line renders on every page, including "Not published", so a missing line can't be read as reassurance. Sources are as described in the Lifestyle amenities, Commute and healthcare notes: hospital locations are OpenStreetMap features tagged with an emergency department (© OpenStreetMap contributors, ODbL).