A new approach to property valuation
Many of today’s AVMs use data-driven linear algorithms specific to a region, metro area, or neighborhood. At GeoPhy, we take a different approach. We use a wider breadth of data to identify global patterns and integrate hyperlocal features to refine analysis specific to a property. This key innovation allows GeoPhy to explore non-linear relationships that expose relevant features that drive property value.
Initially launched in the Netherlands as OfficeRank, we started as data aggregators, providing offices with a quality score based on contextual data.
With a vast database of enriched property data and machine learning expertise, we changed our name to focus on leveraging our core competencies to uncover property features that drive value.
Approached to help a large U.S. government-sponsored enterprise (GSE) develop faster, more accurate independent property valuations for its multifamily loan books worth in excess of US$280bn.
Our efforts result in the first machine learning-enabled valuation solution proven for U.S. multifamily properties. An initial 9% median absolute error (MdAPE) compared favorably against the 10% – 15% rate of traditional appraisals.
We’ve grown the team, enhanced our product (5.85% MdAPE with ongoing refinement), and formed new business partnerships to bring you even better and faster property values.
We introduced EVRA, the commercial real estate industry’s first valuation platform that goes beyond providing a simple valuation by exposing the underlying drivers of value for commercial multifamily properties in the US. Meanwhile in the Netherlands, we introduced GeoPhy WOZ, a residential AVM that provides accurate valuations of the entire Netherlands housing stock for property tax assessments.
Our data science and technology analytics powers the launch of Apprise, a joint venture with multifamily underwriting leader Walker & Dunlop to provide a more objective, consistent, and transparent FIRREA-compliant multifamily appraisal delivered in as little as five days. We also launched GeoPhy Neighborhoods so you can analyze properties’ surroundings more granularly than at the submarket level.
Over the last five years, we’ve grown from a small team of engineers, economists and data scientists working out of a garage (yes, that’s true), to a global employer with staff across The Netherlands, the United States, the United Kingdom, and Lithuania.
Working with GeoPhy to build Apprise by Walker & Dunlop gave us the powerful data science and analytics technology necessary to revolutionize the appraisal process. We will now be able to better serve borrowers and lenders alike.
GeoPhy is instrumental in the data strategy for our global real estate portfolio.
GeoPhy is able to provide us with unique and innovative data and insights into our investment portfolio.
GeoPhy’s hyperlocal data reduces information asymmetry between local managers and centralized investors, which helps us prevent strategy drifts and minimize portfolio risk.
GeoPhy helps us better leverage our data and accurately track and value our nationwide loan book.