Discussions

Hi there,

I am curious to know how successful the merge of property and consumer data is supposed to be over the ADDRID field. There are about 84 million property records and 250 million consumer records in the Verisk datasets - after performing an inner join it seems that not every property record has a match with the consumer data and many more consumer records don't have a property match. Is this because the consumer dataset is of a higher quality and there are simply more records the property dataset doesn't have? I was estimating that each property record has about 2.5 consumer records attached, in which case the number of properties required in the dataset to match all the consumers would need to be higher.

When I filtered the consumer data to only the geographies of interest, about 75% of those consumer records had a property match. Let me know if this sounds reasonable.

Davis