this is property 806 chatGPT advice, real investor returns & what drives capital growth with Luke Metcalfe
Feb 17, 2026
Luke Metcalfe, founder and data scientist at Microburbs who models suburb-level property data. He discusses whether AI like ChatGPT can pick hotspots and where it fails. Short takes on machine strengths versus human nuance. How Microburbs finds suburbs and street-level winners. Why top investors vastly outperform medians, the value of views and scarcity, unit supply risk, NIMBY signals and migration’s sweet spot for growth.
AI Snips
Chapters
Books
Transcript
Episode notes
ChatGPT Isn’t A Property Forecaster
- Large language models like ChatGPT predict the next word and craft plausible text, not reliable property forecasts.
- Luke Metcalfe warns they often average positive sales collateral and can mislead property investors.
Set Strategy Before Scanning Suburbs
- Define your investment strategy before scanning suburbs for buys.
- Use tools like Microburbs’ suburb and property finders only after setting growth, yield and price targets.
Real Investor Returns Vary Wildly
- Investor outcomes vary hugely: median after-cost return ~4.5% p.a., top 10% ~17% p.a., bottom 10% lose heavily.
- Being in the top decile requires selection and skill, not broad market exposure.
