Article | Intelligent Investment
To better uphold consumer privacy in our analytic practices, CBRE | Forum Analytics has implemented a randomization technique to protect customer address data used in our tools while still allowing for accuracy in our modeling and analytics.
Protect and Randomize Your Data
The CBRE | Forum Analytics Randomization Protocol (Protocol) is a proprietary algorithm which adjusts all data points in a dataset within a random distance, appropriate to the surrounding population density. For example, in a census block group with more than 50 residents or workers, the Protocol will randomly move data points within that block group. For areas with less than 50 individuals, or an area larger than 40 square kilometers, the protocol will relocate data points between 45 meters and 1.2 kilometers in a random direction. As data points may represent a consumer’s household, applying the Protocol means that data points used in our platform are unlikely to correlate to a household or a consumer in the underlying data provided by our clients, making re-identification less likely.
Testing the Data
Psychographic analysis relies on precision in customer mapping to block groups in order to generate consumer insights, and our rigorous testing of the Protocol found no meaningful change in the top-rated segments or more likely customers. The result is that conclusions drawn from our analytical processes were not materially impacted by application of the Protocol.
When reviewing application of the Protocol to legacy data, our testing found that less than 10% of customers were assigned to a different census block group, moving them into a different demographic category. This made very little impact analytically. When looking at psychographic analyses we saw a change of less than 0.25% in the percentage of the customers represented by the top 10 segments – well within a normal margin of error.
We are confident in the results that found no material changes to impact the required accuracy. At CBRE | Forum Analytics, we believe the random relocation of consumer household data will not impact our ongoing analytics, while protecting our clients’ customer privacy.
More about Data Privacy at CBRE
With the broadening of definitions of “personal information” under new privacy laws like the California Consumer Privacy Act, randomization techniques such as those applied in the Protocol help to protect your data. CBRE’s Global Data Privacy Office manages a global data privacy program to advise on evolving global privacy laws and help support CBRE businesses to comply with applicable law.