The Cube V6 experiment

Humans make predictions all day, usually without noticing the ingredients. Mine include changing lanes just before the lane I left starts moving. Clearly, there is room for improvement. The Cube experiment asked what happens when you track property-level facts and wider market conditions through time, across tens of thousands of parcels, instead of pulling a few comps when somebody finally asks a question. Signal Intelligence organized that history and tested whether it could rank likely sellers using only information available at the time. Historical tests found a useful concentration of subsequent transactions near the top of the list, with results that depended on the sample, time period, and market conditions. From a distance, it looks like the three of hearts turning up in your pocket. Up close, it is an enormous amount of dated evidence, testing, and correction. The practical idea is to triage the prospect list and show up understanding the forces around a property. V6 is a research experiment, not a promise of a sale date or an owner’s intentions. It helps me choose a better conversation. I still have to make the call.