Case Study: homes.co.nz achieves up to 9.2% lower AVM error with Restb.ai property condition scores

homes.co.nz reduces AVM error by 9.2% with Restb.ai

The customer homes.co.nz, a New Zealand property portal, sought to improve the accuracy of its proprietary automated valuation model (AVM). Their challenge was that most AVMs have a blind spot regarding a home's interior condition, which limited their valuation precision. They partnered with vendor Restb.ai to explore its property condition score technology.

Restb.ai implemented its computer vision technology to analyze images of kitchens and bathrooms, providing condition scores in less than a second. This solution allowed homes.co.nz to account for interior quality within their AVM. The results showed a significant improvement, with the AVM's median error rate dropping by up to 9.2% in certain regions, leading to more accurate and transparent home valuations.

homes.co.nz