How We Calculate

Liveability Rating (A+ / A / B / C / D)

The headline grade on each suburb page combines five state-percentile-ranked indicators into a single 0–100 composite, then bands the result into letters. Ranking is within-state so a suburb is compared to honest peers, not to all of Australia. Suburbs with population < 500 stay Unrated. The letter grade is ours, not an official assessment — no agency grades Australian suburbs.

Component weights

• Schools quality (25%) — % of nearby schools rated above similar (requires ≥ 3 schools)
• Safety (25%) — average of fire, crime, and flood risks (inverted)
• SEIFA IRSAD (25%) — ABS socio-economic advantage percentile
• Walkability (15%) — Walkability Score within state
• CBD access (10%) — driving distance to state capital, inverted

Missing indicators (e.g. no schools data, no crime data in TAS/NT) are dropped and the remaining weights re-balanced — there is no fixed penalty for unmapped inputs.

Letter bands

Bushfire Risk

Based on historical bushfire boundary data from national and state fire agencies. For each suburb, we calculate the overlap ratio — total burnt area divided by the suburb's size — to measure fire exposure relative to area. Suburbs are then percentile-ranked among all suburbs with any fire history, mapped to a 10–100 scale. Any suburb with recorded bushfire overlap gets a minimum score of 10. A score of 0 means no recorded bushfire history.

Crime Risk

Based on the most recent year of officially recorded offences from state crime statistics agencies. Each offence category is weighted by severity — violent crimes (assault, robbery, homicide) carry higher weight than regulatory offences. The weighted total is divided by suburb population to produce a per-capita rate, then percentile-ranked within the same state for a fair comparison across different data sources. Any suburb with crime data gets a minimum score of 10. Crime data is available for VIC, NSW, SA, QLD, WA, and ACT. Tasmania and Northern Territory are not yet supported as suburb-level data is not published by their police agencies. A score of 0 means no crime data is available.

Transport Access Score

Measures how well-connected a property is to public transport. The score combines three factors: mode diversity (40%) — how many different transport types are nearby; proximity (35%) — how close the nearest stop is; and stop density (25%) — total number of stops within 1 km. Rail-based modes (train, metro, tram) receive a bonus for their higher capacity and frequency. A score of 0 means no public transport stops were found within 1 km.

Walkability Score

A 0–100 score that summarises how walkable a suburb or property is, combining the Transport Access Score with the density of everyday amenities nearby. It's computed from the same data on both suburb and property pages. Higher scores mean more daily errands can be done on foot.

Component weights

• Public Transport (30%) — the Transport Access Score above, within 1 km
• Schools (15%) — ACARA school locations within 1.5 km
• Parks (12%) — parks & reserves within 800 m
• Supermarkets (12%) — supermarkets within 1 km
• GPs (10%) — medical centres within 1 km
• Pharmacies (8%) — within 1 km
• Childcare (8%) — ACECQA-approved centres within 1 km
• Hospitals (5%) — within 5 km (drive range, not walk)

Fuel stations are excluded, because people rarely reach them on foot. They still appear in the suburb-level Amenities section as a standalone count.

How each component is scored

For each amenity category we count how many are within its radius and apply a saturating curve: 0 amenities → 0, 1 → 60, 2 → 80, 3 or more → 100. Presence matters more than abundance — the first GP nearby is a big jump, the second smaller, and the third plateaus. The final score is the weighted average of all components, rounded to the nearest integer.

School Proximity

On property pages, shows the nearest school to the address grouped by school type (primary, secondary, combined) and by sector (government, catholic, independent). Distance is straight-line from the property; walking-time estimates show for schools within 1.2 km. On suburb pages, shows every school whose location falls inside the suburb boundary, grouped by type. Note: schools are based on distance and boundary containment, not official school zones — check with your state education department for enrolment catchments.

School Quality Band

Each school gets a peer-relative label — Above, At, or Below similar schools — instead of a raw score. The peer group is defined by objective inputs, not outcomes, so schools compete against realistic peers rather than the national average.

How the band is computed

1. For each school with at least one year of NAPLAN data, we take the most recent year and compute an unweighted mean across every tested year level (Y3, 5, 7, 9) and all five domains (Reading, Writing, Spelling, Grammar, Numeracy).
2. The school's peer group = schools in the same state, of the same type (primary / secondary / combined), with ICSEA within ±50 (widened to ±100 if fewer than 20 peers are found).
3. Within the peer group, the composite score is ranked. Top third → Above similar, middle third → At similar, bottom third → Below similar.
4. Schools without ICSEA, without NAPLAN data (early childhood, senior secondary colleges, environmental education centres), or in peer groups with fewer than 20 members get no band — the pill is hidden.

What the band does not tell you

• Absolute performance — a school in a wealthy area rated At similar schools may still score higher in raw terms than a school in a disadvantaged area rated Above similar schools.
• Individual-student outcomes. NAPLAN reports averages, not how any one child will progress.
• Whether the school fits your family. Enrolment policy, curriculum focus, facilities, and culture are not captured.

Per-school trend arrow

Each school also shows a tiny ↑ / → / ↓ arrow next to its quality-band pill. The arrow compares the school's composite NAPLAN mean between the earliest and latest year it has data for (up to a 3-year window). A delta above +20 is improving, below −20 is declining, otherwise stable. Schools with fewer than 2 years of data get no arrow.

Suburb Schools Score (0–100)

On each suburb page we aggregate the individual school bands into a single 0–100 score. The formula uses additive Bayesian smoothing so a single above-similar school in a rural suburb doesn't hit a misleading 100.

score = round(
  (above_count * 1.0 + at_count * 0.5 + PRIOR * 0.5)
  / (above_count + at_count + below_count + PRIOR)
  * 100
)

PRIOR = 3

The prior is equivalent to adding 3 phantom "at similar" schools to every suburb's denominator. It pulls small-sample scores toward 50 — 1 above-similar school scores around 62, not 100 — and has negligible effect once a suburb has 10+ rated schools. Unrated schools (no ICSEA, no NAPLAN, or too-small peer group) are excluded entirely. If a suburb has zero rated schools, the score is hidden rather than shown as 0.

SuburbScan derives the peer-relative band from ACARA's published data. Each school's ICSEA value and percentile are shown alongside the band; per-domain NAPLAN scores by year level are not displayed — refer to ACARA directly for those.

Gross Rental Yield

Annual rent as a percentage of purchase price: median weekly advertised rent × 52 ÷ median asking price × 100. Houses and units are calculated separately and never blended — unit yields run systematically higher than house yields in the same suburb, so a combined figure would flatter houses and mislead.

What "gross" means

Gross yield is the headline figure before any costs. It does not deduct strata or body-corporate fees, council rates, water, landlord insurance, property management (typically 5–8% of rent), maintenance, or vacancy between tenancies. Net yield after those costs is commonly 1–1.5 percentage points lower, and varies far more by individual property than by suburb. Treat the figure as a suburb-level benchmark for comparing areas, not as a projected return on a specific purchase.

Why asking prices, not sale prices

The denominator is the median advertised asking price from public listings over the last 6 months — the same source, time window and suburb boundary as the rent figure above it, so the two sides of the ratio are directly comparable. Valuer-General settled-sale data would be more authoritative but is only available to us in NSW, SA and VIC, and covers a different period. Because asking prices sit above eventual sale prices, this method understates yield slightly; we prefer the conservative direction.

The property-mix caveat

The homes advertised for rent in a suburb are not the same homes advertised for sale. Rental stock skews smaller and older in most suburbs, which can push the calculated yield either way. This is the main reason a suburb-level yield should never be read as the return on any individual property.

When we don't show it

Both sides need at least 3 listings in the 6-month window before we publish a median, so a yield only appears where both cleared that floor. Where the weaker side has fewer than 8, the figure is labelled as a small sample. Results outside 0.5–20% are treated as data errors and suppressed rather than shown.

SuburbScan is not a financial adviser and this figure is not investment advice.

Council Rates

The figure on each suburb and property page is the LGA-wide average residential rate published by the relevant state portal, not an estimate for your specific property. It changes by council boundary, not by suburb. Year-on-year change compares the most recent published FY against the prior FY.

What's included

We show the general residential rate only — what you're assessed via your council's rate-in-the-dollar applied to your property valuation. We do not bundle in utility charges (water, sewerage, garbage) or state-imposed levies (emergency services, NRM). Other comparison sites that report a higher figure are typically including those add-ons. Your actual rate notice will be higher than the figure shown here for that reason.

Coverage

• NSW — Office of Local Government Time-Series xlsxs. Average Residential Rate per assessment.
• VIC — Local Government Performance Reporting Framework (LGPRF). Average rate per property assessment.
• QLD — DLGWV Comparative Information Report. Average General Rate payable per residential property, averaged across listed major urban centres in each council.
• WA — Department of Local Government MyCouncil OData feed. Approximated from total rate revenue ÷ residential dwellings (see WA caveat below).
• SA — Local Government Grants Commission Database Reports (PDF). Total Residential Rates per Rateable Residential Property (Report 6).
• TAS — LIST Tasmanian local government Consolidated Data Collection. Approximated from total residential rates revenue ÷ census dwellings per LGA.
• ACT — ACT Treasury's annual Aggregate Land Value for General Rates Settings paper, published with each Budget. ACT is a single rating jurisdiction (no councils — the territory government performs council functions itself), so every ACT suburb shows the same figure. We read the average general rate per residential property directly from that paper, blending unit and non-unit properties by the property counts it gives. UV basis (3-year average unimproved value). The FESL, Safer Families Levy, and health levy are excluded for cross-state methodology consistency. Until August 2026 we estimated this figure instead, by querying the Revenue Office calculator API at one representative valuation. That understated it: the ACT schedule is progressive, and units — around two in five residential properties — are rated well below houses, so no single valuation can stand for the average.
• NT — per-council Declarations of Rates and Charges (each council's annual legal notice). No centralised dataset; 16 of 18 NT councils ingested by manually transcribing residential rate parameters into config. UV basis across all NT councils. Most NT regional councils use flat per-allotment charges rather than rate-in-the-dollar formulas. Coomalie and Groote Archipelago not currently shown — Coomalie hasn't published a recent declaration; Groote is a 2023 council split that the 2024 ABS LGA vintage doesn't reflect yet.

Valuation methods by state

The single biggest reason direct dollar comparisons between states can be misleading. A $500,000 property in NSW and a $500,000 property in VIC are measured on entirely different scales.

• Unimproved Value (UV) — used in NSW, QLD, ACT. Land only, ignoring buildings or improvements. A vacant block and a house on an identical block next door have the same UV. UV states tend to apply a higher rate-in-the-dollar because the valuation base is smaller.
• Capital Improved Value (CIV) — used in VIC. Land + buildings + permanent improvements. Essentially the market value of the whole property. VIC councils typically apply a lower rate-in-the-dollar than UV states to raise comparable revenue.
• Gross Rental Value (GRV) — used in WA for most urban properties. Estimated annual rental income. Rural and mining properties in WA use UV instead — this dual system is why some remote WA shires show unusually high average rates.
• Capital Value (CV) — used in SA (most councils) and parts of TAS. Similar to VIC's CIV. Some SA councils use Site Value (SV, equivalent to UV) or Annual Value (AV, rental-based) instead — the basis is shown on each suburb's rates card. SA councils also apply a Natural Resources Management (NRM) levy on top of general rates.
• Assessed Annual Value (AAV) — used by some TAS councils (notably Hobart, Glenorchy, Launceston). Rental-based, similar to WA's GRV. Other TAS councils use CV, making TAS the most heterogeneous state to compare within.

Rate caps and pegging

NSW rate-pegs annual general-rate revenue increases via IPART (typically 3–4% per year). VIC caps annual general-rate-revenue increases via the Minister for Local Government (recently 2–3%). These caps mean year-on-year change is small and uniform across their councils — multi-year cumulative change is more informative than a single FY.

WA caveat

WA councils don't publish a residential-only rate. The figure shown is council-wide rate revenue divided by residential dwellings, so shires with significant mining or commercial revenue can show inflated numbers. A remote WA shire showing $20k+ per year is real published data — most of that revenue comes from iron ore tenements, not residences.

What council rates don't include

• Waste collection & recycling charges (often a separate fee)
• Special rates and levies for specific projects
• Emergency services levies (varies by state)
• State-imposed fees rolled into rate notices

Always check your council's rate notice or contact them directly for your exact bill.

Aircraft Noise Risk

Based on ANEF (Australian Noise Exposure Forecast) contour data from Airservices Australia and state planning authorities. For each suburb, we check whether its boundary intersects any mapped ANEF noise contour. The highest ANEF level found determines the noise exposure. Suburbs are then scored based on the severity of the highest contour — ANEF 20 is the lowest mapped level (noticeable noise), while ANEF 40+ is considered unacceptable for residential use. A score of 0 means no ANEF noise contours intersect the suburb. Property pages check whether the specific address point falls within a noise contour for precise assessment.

Coastal Erosion Risk

Based on DEA Coastlines satellite data (1988–2024) from Geoscience Australia. For each suburb, we query the nearest high-certainty erosion monitoring point within 5 km and extract the worst (most negative) shoreline change rate. Only negative rates count as risk — positive rates (accretion) are ignored. Suburbs are then percentile-ranked among all suburbs with erosion data, mapped to a 10–100 scale. Any suburb with nearby erosion gets a minimum score of 10. A score of 0 means no monitored coastline within range. Property pages show individual monitoring points within 1 km with detailed rate and certainty data.

CBD Distance

Real driving distance and travel time from the suburb centroid to the state capital CBD, computed once per suburb via the Mapbox Directions API and cached in the database. The landmark used as the CBD reference point per state is:

• NSW — Sydney Town Hall
• VIC — Flinders Street Station
• QLD — Queen Street Mall
• WA — Perth CBD
• SA — Victoria Square (Adelaide)
• TAS — Hobart waterfront
• ACT — Civic (Canberra)
• NT — Darwin Mall

Distances are for the centroid of the suburb, not any individual address, so expect variation within larger suburbs. Drive times reflect typical traffic conditions at the time Mapbox routed the query; they're a guideline rather than a peak-hour commute estimate.

Land Area and House Blocks

A property's land area comes from the state's cadastre: the surveyed boundary of the lot under the address. We calculate the area from that boundary, because most older lots do not record one. If the lot holds more than one street address, as a block of units or a dual occupancy does, the page shows a shared block rather than one home's land.

A house block is a lot that is not a strata or unit lot and holds exactly one street address. We compare a property with the middle half of its suburb's house blocks: below that range is "smaller than most", above it is "larger than most". The comparison is shown only where a suburb has at least 30 house blocks. Lot boundaries are available in New South Wales, Victoria, Queensland, South Australia, Tasmania and the ACT.

Growth Areas and New Homes Planned

Growth precincts come from the Victorian Planning Authority, the NSW Department of Planning and Economic Development Queensland. A property page lists the precinct the address is in and any precinct within 3 km, measured to the precinct's nearest edge. Each agency lists only the plans it leads, so no precinct does not mean no plan.

In Queensland, new homes planned come from the Queensland Government's residential land supply study, which gives each parcel an expected timing and a dwelling yield. The yield is published as a range, such as 21 to 50 dwellings, so we add up the low and high ends separately. A parcel listed as over 500 dwellings has no upper limit, so a total that includes one reads "at least". A property page counts land within 2 km of the address; a suburb page counts land in or crossing the suburb.

Mortgage Brokers

We count the people on ASIC's credit representative register who work with one of the national aggregators listed by the Mortgage and Finance Association of Australia, once each, at their registered suburb. Brokers with their own credit licence or another aggregator are not counted, so the figure is likely to be lower than the true number. A place holding at least 50 of one aggregator's brokers and at least 5% of them is treated as a head office and left out.

Ranked suburb lists

The ranked lists — "best suburbs for first-home buyers", "where the Greek community lives", "most educated suburbs", and so on — all score the same candidate pool: suburbs within the capital's metro radius, above a population floor, that we consider indexable. Every ranking is computed within that one capital, so a suburb is measured against its own metro rather than the whole country. Combinations with too few qualifying suburbs aren't published.

Livability rankings

Lists like first-home buyers, rental yield, public transport, NBN connectivity, EV adoption, investor / owner-occupier and young professionals blend several signals into one score — each signal ranked as a percentile within the city's pool, then combined with published weights (shown in the "How we ranked these" note on each page). Hard eligibility gates (minimum listings, a safety floor, etc.) keep unsuitable suburbs out before scoring.

Census demographic rankings

The community and demographic lists come straight from the 2021 Census — each is a share of a suburb's residents, ranked within the capital, among suburbs with at least 2,000 residents:

• Religion — share of residents who reported a given religion.
• Age cohorts — share aged 0–14 (young families) or 65+ (seniors).
• Language — share who speak a language at home. We use the full "language used at home" count, not the first-or-only-language subset, which undercounts by roughly half.
• Most educated — share of residents aged 15 and over with a bachelor degree or higher, measured against the 15+ base (not total population, which would dilute the figure with children).
• Highest income — median total household income per week (a value, ranked directly — the median resists skew from a few very high earners).
• Where communities live — ranked by the most reliable single signal (ancestry for most communities; country of birth for the Indian community, whose ancestry splits across many sub-groups), while every row shows all three angles — ancestry, country of birth, and language — side by side.

Small-cell floors and the 2,000-resident gate keep census noise out. These are descriptive snapshots of who lives where, drawn from public ABS data — not a judgement about, or a recommendation to seek out or avoid, any suburb.

Suburb Finder

The Suburb Finder turns the ranking engine around: instead of a fixed list, you set how much six factors matter — affordability, safety, public transport, walkability, schools, and closeness to the city — from "ignore" to "must-have", plus a property type and budget. We then rank the chosen capital's suburbs to match.

How the match score works

Each factor is ranked as a percentile within the city's pool, then blended using your importance weights into a single 0–100 match score. Budget is a hard filter — suburbs outside your price range are removed before scoring. Where a suburb is missing data for a factor, that factor is treated as neutral, so a gap neither helps nor hurts it unfairly.

It's fully deterministic: the same inputs always produce the same ranking. No AI or language model decides which suburbs appear — the score is a transparent, reproducible blend of the same public data behind the rest of the site.