H3 Spatial Analytics: Build Property Heatmaps with HtAG

See variation that a suburb median can hide

For the resolution-5–9 scale comparison, complete endpoint directory and geography crosswalks, read our H3 format, endpoints and geography-mapping guide.

A suburb summary is useful, but it compresses every part of the suburb into one number. H3 adds a consistent hexagonal grid so analysts can inspect how a supported measure varies across smaller geographic cells.

The map below shows a real, bounded HtAG snapshot for Richmond VIC 3121. It contains five H3 resolution 8 cells for houses across all bedroom and bathroom counts. Select a cell to inspect its median weekly rent and use the street map to understand where the aggregate sits.

View the demo screenshot

What this Richmond map says

The five captured cells range from $874 to $945 per week—a $71 spread. An agent or property manager could use this view to choose where to investigate comparable rental evidence before preparing a customer briefing. Cell 88be635621fffff has the highest median weekly rent in this set at $945, while cell 88be63562dfffff has the lowest at $874. The snapshot was captured on 10 September 2026.

Each value is a geographic aggregate. It is not the rent of a specific property, and the differences do not identify a preferred street or neighbourhood. Property condition, size, bedrooms, precise location and other attributes still matter when assessing an individual home.

How H3 helps spatial analysis

Administrative boundaries are useful for reporting, but they vary in size and can place similar streets into different suburb buckets. H3 provides a hierarchical grid with stable cell identifiers. Analysts can use a coarser resolution for a broad view and a finer supported resolution when the question needs more local detail.

HtAG’s current spatial datasets support resolutions 5 through 9. Resolution choice affects geographic meaning, row count and cost. A fine suburb view should still use enough observations per cell to support the measure, while a coarse cell can extend well beyond the locality used to select it.

One endpoint, one measure

The map’s values are H3-cell median weekly rents for the stated resolution and filters, taken from the API response retrieved 10 September 2026. They are cell aggregates rather than the suburb-level Median Rent. Risk layers also require careful direction: a public H3 flood or bushfire value expressed as an exposure percentage rises as exposure rises, while HtAG’s suburb risk indices run from 0 to 100 with higher scores meaning lower relative risk.

The current H3 API separates measures into explicit operations. Median price, median weekly rent, gross yield, days on market and flood exposure are requested independently. A metric response identifies the measure and unit, then returns cell rows with an h3_index, value and observation count.

For this map, the data request uses the median-rent operation with one geographic filter:

GET /v1/spatial/h3/price-rent/median-rent
  ?resolution=8
  &loc_pid=VIC2167
  &property_type=house
  &beds=ALL
  &baths=ALL
  &min_n=10
  &strict_granularity=true

Use exactly one of loc_pid, lga_pid, sa2_code or gcc_name in a metric request. Keeping a single explicit area filter makes the map extent and interpretation easier to audit.

A compact production workflow

1. Resolve Richmond to the locality ID VIC2167.
2. Request median weekly rent at resolution 8 with a minimum observation threshold.
3. Retrieve the boundary for each returned H3 identifier.
4. Join values to their matching polygons without dropping or inventing cells.
5. Render the polygons over a geographic basemap with roads, labels and attribution.
6. Publish the captured data and geometry together so the browser makes no HtAG API request.

Build a map people can read

Use a sequential colour scale for rent, show the actual minimum and maximum in the legend, and display the exact dollar value on hover, tap and keyboard focus. Keep the hexagons partly transparent so roads and place labels remain visible. The basemap supplies geographic context; it is separate from the HtAG metric snapshot and must retain its provider attribution.

When adding another layer, request and validate that measure independently. Do not imply that every cell contains price, rent, yield and risk in one record. Keep the property filters, resolution and selected cell cohort consistent before comparing layers.

Read the median-rent H3 reference, review the supported H3 resolutions, or sign in to the developer portal to test a bounded area.