"Big data" is receiving considerable attention lately. Having recently given a talk on programming using big data (slides), it strikes me that I should write something here on big data as it applies to commercial real estate (CRE). It is often said that big data is data too large to fit in a spreadsheet, i.e., it is voluminous. For CRE, this might not be the most useful analogy. Another feature of big data is variety—banging together spreadsheets that weren't designed to talk to each other. For CRE, I think this is the better analogy.
CBRE has a lot of proprietary data sitting in data "puddles." Big data is creating a data "lake" from those puddles by linking data feeds, and then using the power of machine learning to derive insights for our clients. Our just-introduced Live, Work, Play (LWP) Index is an example of a big data exercise in CRE. We've linked a number of proprietary data sources on key CRE drivers to map where the people are (Live), where the office space is and how much of it there is (Work), and where the retail space is and how much of it there is (Play). We then use a sophisticated machine learning algorithm to generate an index that can be used for site selection. The higher the index, the stronger is the LWP connection and potentially the better the performance.
We will be adding additional features, such as rent performance, to improve the predictive analytics of the LWP Index. We are just getting started. Stay tuned.