Where Does Picki Get Its Data? A Complete Guide to Our Property Data Sources
When you're making a property investment decision worth hundreds of thousands of dollars, you need to trust the data behind it. At Picki, we believe data transparency isn't just good practice — it's essential. That's why we're pulling back the curtain on exactly where our data comes from, how often it's updated, and what metrics each source powers.
Picki analyses over 2 million Australian properties using data from government sources, industry databases, and our own proprietary data pipeline. Here's the complete breakdown.
1. ABS Census Data — Demographics & Population
The Australian Bureau of Statistics (ABS) Census is the backbone of our demographic intelligence. Conducted every five years (most recently in 2021), the Census provides granular data at the suburb (SSC), SA1, SA2, and LGA geographic levels.
What we extract:
- Population figures — total population by suburb and LGA, with historical comparison across Census periods (2011, 2016, 2021)
- Dwelling counts — total dwellings, new dwelling approvals, and housing supply trends using GNAF (Geocoded National Address File) data
- Median household income — income distribution by suburb and LGA, tracked across Census years to calculate income growth trajectories
- Tenure type breakdown — owner-occupier vs renter ratios, public housing intensity, and private rental proportions at the SA1 level
- Property type distribution — breakdown of houses, units, townhouses, and other dwelling types per suburb
- Employment and industry data — worker industry breakdown at the SA2 level, commute type data, and employment diversity scores at the LGA level
How often it's updated:
Census data refreshes every five years, but we supplement it with annual ABS estimates for population projections and income growth between Census periods. Our LGA population projections use three-year forward estimates.
What it powers in Picki:
Demographics cards on suburb pages, population growth charts, income comparisons, owner-occupier ratios, public housing intensity metrics, and several components of the R-Score.
2. State & Territory Valuer General Data — Property Sales
Every state and territory in Australia maintains a Valuer General (or equivalent) that records all property transactions. This is the gold standard for actual sale prices — not asking prices, not estimates, but what properties actually sold for.
What we extract:
- Settled sale prices — every residential property transaction across Australia
- Rolling median prices — calculated on 90-day and 365-day rolling windows at suburb, LGA, and state levels
- Sales volume — the number of transactions per period, a key indicator of market liquidity and confidence
- Price deltas — suburb-level price changes compared against LGA and state benchmarks to identify outperformers and underperformers
How often it's updated:
New sales data flows through as settlements are registered — typically monthly, with a slight lag of 4–8 weeks from settlement to registration depending on the state. We process and bake this data into our metrics on a rolling monthly cycle.
What it powers in Picki:
Median price displays, price growth charts, sales volume indicators, suburb research metrics, property appraisal comparable sales, and several R-Score components.
3. Rental Listing Data — Vacancy Rates & Yields
Understanding the rental market is critical for investment analysis. We track rental listings across major Australian property portals to build a real-time picture of rental supply and demand.
What we extract:
- Active rental listings — the number of properties currently advertised for rent in each suburb
- Vacancy rates — calculated by dividing active rental listings by the total rental dwelling stock (using Census tenure data). We calculate two versions: a standard rate and an industry-standard rate that only counts listings on market for 21+ days, aligning with SQM Research methodology
- Median weekly rent — rolling median rental asking prices by suburb, tracked over 90-day and 365-day windows
- Rental yield — gross rental yield calculated from the median rent relative to median sale price, tracked as rolling metrics
- Rent price changes — delta tracking of rental price movements over time
How often it's updated:
Rental listing data is ingested and processed daily. Our vacancy rate and yield calculations refresh monthly as part of the data bake cycle.
What it powers in Picki:
Vacancy rate displays, rental yield calculations, rent price trend charts, rental income estimates, and the vacancy/yield components of the R-Score.
4. Property Listing Data — Days on Market & Demand Signals
For-sale listing data tells us how quickly the market is moving and whether sellers or buyers have the upper hand.
What we extract:
- Days on market (DOM) — how long properties take to sell from listing to sale, averaged across the suburb on a rolling basis
- Listing volumes — the number of properties actively listed for sale, indicating supply pressure
- Price changes during campaigns — tracking when agents increase, decrease, add, or remove asking prices (a signal of vendor motivation)
- Vendor discounting — the gap between initial asking price and final sold price, measured as a percentage
- Days on site (DOS) — listing duration metrics including delta tracking over time
How often it's updated:
Listing data is monitored daily. Our baked metrics (rolling averages, DOM calculations) are processed monthly. Price change alerts for shortlisted properties are sent in real-time.
What it powers in Picki:
Days on market displays, vendor discounting metrics, campaign stress indicators, property shortlist notifications, and the DOM/DOS components of the R-Score.
5. GNAF — Australia's Geocoded Address Database
The Geocoded National Address File (GNAF) is Australia's authoritative database of every physical address. Maintained by Geoscape Australia (formerly PSMA), it's the foundation that lets us link properties to their geographic identifiers.
What we extract:
- Address-to-location mapping — linking every property to its suburb, SA1, LGA, and state boundaries
- New dwelling supply tracking — using address creation dates to measure new housing supply over time
- Property identification — unique GNAF IDs that allow us to track individual properties across multiple data sources
How often it's updated:
GNAF releases quarterly updates. We integrate each release to capture new addresses and boundary changes.
What it powers in Picki:
Property search, address matching, supply trend charts, new dwelling counts, and the geographic backbone of every metric we calculate.
6. Infrastructure & Development Data — LGA Project Spend
Infrastructure investment is a leading indicator of future growth. We track government and private development project data at the LGA level.
What we extract:
- Project spend per capita — total infrastructure and development expenditure divided by LGA population, providing a normalised comparison across regions
- Project distribution — breakdown of where development dollars are flowing, stored as percentile rankings in our data warehouse
- Proximity to utilities — LGA-level scoring of access to essential infrastructure and services
How often it's updated:
Infrastructure data is refreshed quarterly as new project announcements and completions are recorded.
What it powers in Picki:
Infrastructure spend indicators on LGA profiles, utility proximity scores, and the project spend per capita component of the R-Score.
7. AI-Powered Property Appraisals — Nightly AVM Processing
Picki runs an advanced AI-powered Automated Valuation Model (AVM). Every night, our system processes comparable sales data, market trends, and property characteristics to generate up-to-date property appraisals.
What it does:
- Comparable sales analysis — identifies and weighs recent sales of similar properties in the surrounding area
- Market condition adjustment — factors in current suburb-level demand, supply, and price trends
- Property-specific scoring — accounts for bedrooms, bathrooms, land size, property type, and location adjustments at the SA1 micro-level
- Portfolio monitoring — for registered users, tracks appraised values across your entire portfolio and sends alerts when meaningful changes occur (1%+ movement)
How often it's updated:
Property appraisals run nightly. You literally wake up to fresh valuations.
What it powers in Picki:
Property value estimates, portfolio value tracking, appraisal change notifications, and the AI appraisal cards on property pages.
8. Socioeconomic & Liveability Indices
Beyond pure property metrics, we incorporate broader socioeconomic data that influences long-term property performance.
What we track:
- SEIFA scores — the ABS Socio-Economic Indexes for Areas, measuring relative advantage and disadvantage
- Income growth trajectories — median household income growth between Census periods at suburb and LGA levels
- Employment diversity — how diversified the local employment base is at the LGA level (a resilience indicator)
- Population growth projections — three-year forward population estimates at the LGA level
What it powers in Picki:
Socioeconomic indicators on suburb profiles, liveability context, and several R-Score components including the population growth vs supply analysis.
How the R-Score Brings It All Together
The R-Score is Picki's proprietary composite investment score. It's not based on a single metric — it's a weighted calculation across 26 individual data points, each drawn from the sources described above. Here's how it works:
Each suburb receives a raw value and a percentile ranking for every metric. The R-Score then applies conditional weighting — a metric only contributes to the score if it falls within a healthy range (for example, vacancy rate only counts if it's below a threshold, and rental yield only counts if it's above a minimum).
The top-weighted R-Score components:
- Sold price delta — recent price growth momentum at the suburb level
- Rental yield — gross rental yield relative to purchase price
- Median sold price — price point accessibility (favouring entry-level markets)
- Proximity to utilities — infrastructure and services access at the LGA level
- Days on site — market absorption speed
- Price delta vs state — relative performance compared to the state average
- Population projection — three-year forward population growth at LGA level
- State-level price momentum — broader market direction
Additional components include median household income growth, owner-occupier ratio, investment ratio, SEIFA scores, LGA population, rent prices, yield deltas, vacancy rates, public housing intensity, employment diversity, and project spend per capita — each weighted between 2% and 5.7%.
The result is a score out of 100 that reflects the overall investment attractiveness of a suburb, balancing growth potential, yield, demand, and risk factors.
Our Data Pipeline: How It All Comes Together
Raw data is only as useful as the system that processes it. Here's a simplified view of our data pipeline:
- Collection — Data flows in from ABS databases, Valuer General registries, listing portals, GNAF releases, and infrastructure databases
- Processing — Our data engine normalises, cleans, and links data across geographic boundaries (suburb → SA1 → LGA → state)
- Baking — Rolling metrics are calculated (90-day and 365-day windows), percentile rankings are generated across all suburbs nationally, and the R-Score is computed
- Delivery — Processed data is served through our API to power suburb pages, property profiles, portfolio dashboards, and search rankings
This pipeline runs continuously — with nightly property appraisals, daily listing monitoring, and monthly metric recalculations.
Key Takeaways
- Government-grade foundations — Our core data comes from ABS Census, Valuer General sales records, and GNAF — the same sources used by banks and government agencies
- Real-time market monitoring — Rental and sale listing data is tracked daily, giving you current market conditions rather than stale quarterly reports
- 26-metric composite scoring — The R-Score isn't a guess or a vibe — it's a weighted calculation across 26 measurable data points with conditional logic
- Nightly AI appraisals — Property valuations update every night, not once a year like a traditional bank valuation
- Full transparency — We don't hide behind black-box algorithms. Every metric on Picki can be traced back to a verifiable data source
Frequently Asked Questions
How often is Picki's data updated?
It depends on the data type. Listing data and property appraisals update daily/nightly. Rolling suburb metrics (medians, yields, vacancy rates) recalculate monthly. Census-derived demographics update with each ABS Census (every five years), supplemented by annual ABS estimates.
Where does the R-Score come from?
The R-Score is a proprietary composite score calculated from 26 individual metrics spanning sales data, rental data, demographics, infrastructure, and socioeconomic indicators. Each metric is percentile-ranked nationally and weighted according to its predictive importance. Read more about how we use data to find undervalued suburbs.
How accurate are Picki's property appraisals?
Our AI-powered AVM uses comparable sales, market conditions, and property characteristics to generate estimates. Like all automated valuations, they're most accurate in areas with frequent, recent sales of similar properties. We recommend using them as one input in your research alongside professional valuations for major decisions.
Can I see the raw data behind a suburb's score?
Yes. Every suburb page on Picki breaks down the individual metrics that feed into the overall score — from vacancy rates and yields to population growth and income data. We believe in showing our working.
