Precision with Privacy (PwP) from BI Spatial

PwP Defined:

A monthly, spatially anonymized dataset containing the most representative residential points of mobile devices in the United States which can be coupled with various available commercial segmentation schemas.

What we mean by “Precision”

Monthly PwP designations are derived from the calculated month plus the two preceding months of mobile activity and utilizes a recency weighting B I Spatial only includes mobile data observed within residential areas
Each PwP must be within a short distance from a known residential point (household, apartment building or on or near-campus housing building).
PwP is assigned the closest point’s segmentation and the closest point’s Zip+4, with some exceptions.
- Some apartment communities are gated and the Zip+4 lat/longs are inappropriate, typically near the gate. B I Spatial utilizes apartment buildings for more accurate lat/long assignments. These records are then flagged as “apartments” and appended with the nearest corresponding apartment-centric segmentation.
- Devices most often observed in university-sponsored housing are flagged as college students using B I Spatial’s proprietary student housing geofences (18,000 on- or near- campus buildings designated as campus housing by each campus’ student housing department). When this method is used no Zip+4 is assigned.

What we mean by “Privacy”

To protect mobile device users’ anonymity, B I Spatial has created a spatial anonymization process. Although segmentation appending is based on the nearest residence, B I Spatial never provides actual residential lat/longs.

Lat/Long assignments are based on the centroid of each device’s Zip+4. However, in more than 50% of Zip+4s there are fewer than 3 addresses. For those lower count Zip+4s B I Spatial developed a method to anonymize the lat/longs.

Non-Rural vs Rural

- Non-Rural : Within these areas, appending of the nearest, qualifying, neighboring Zip+4’s lat/long and Zip+4 is the best option. These assignments maintain the precision needed for accurate spatial representation and modeling.
- Rural : Within sparsely populated areas more anonymization is required. To increase the level of privacy in these areas, our process identifies the most central Zip+4 for all Zip+4’s within a US Census Block Group. It is important to note that the Block Groups are used for grouping Zip+4s, not appending. That Zip+4’s lat/long and Zip+4 is used for all. Spatially, the derived point is less geographically precise than the non-rural method but is still useful for customer spotting and trade area delineation.

National Mobility Grid (NMG) utilizing H3 Hex Grids

BI Spatial utilizes the NMG for precise analysis.

NMG Defined:

A monthly aggregation of mobile device activity by day and hour for the H3 hex grid cells of the United States.

H3 is a Hexagonal hierarchical geospatial indexing system which allows for easy interpretation of data. Hex overlays remove the burden of comparing data within irregular geographies such as Census Block Groups or postal Zip Codes.

NMG is offered with and without commercial segmentation.

About BI Spatial:

B I Spatial brings 30+ years of expertise in the spatial Business Intelligence arena and nearly a decade within Mobile Analytics arena across a variety of industry segments including Retail, Restaurants (QSR & Full Service), Grocery, Banking, Education, and Hospitality. Learn more here: https://www.bispatial.com/

For questions or to request a custom quote, please contact DMM Sales Dept at (845) 348-7000 x204 or use our contact form.