Zillow Launches New Neural Zestimate, Yielding Major Accuracy Gains
“The new Zestimate algorithm leverages neural networks, the latest machine learning approach, and incorporates deeper history of property data such as sales transactions, tax assessments and public records, in addition to home details such as square footage and location.”
Looks like “neural” is the new “AI”. Count down from 10 until vendors start issuing press releases about their “neural CRM platform”. ????♂️
Wait – so now we know why the Zestimate has always been so far off – for the first time it “incorporates deeper history of property data such as sales transactions, tax assessments and public records, in addition to home details such as square footage and location.”
Well, it’s about time. They should have just used the Cloud CMA methodology – would have been way better from the start.
LOL – just being snarky –
Barton said something profound in the last earnings report. Namely that consumers expect Zillow to iBuy at the Zestimate price – and his goal is to make the Zestimate accurate enough to do that. Great Moon Shot – The commercial AVM market is understood by professional portfolio managers to get you close – but a physical inspection of the house is a critical part of closing in on that last 5% to 10% of accuracy.
I am more surprised that the Zestimate is not using image and video data extraction to get that last mile that a human inspection fills. Or….maybe they are.
“Looks like “neural” is the new AI” – Yep! The Zestimate announcement elicits quite an eye roll. Paraphrasing Greg, a Zestimate can’t smell a cat.