March-2019 Release Notes (v2019-02-28)

Core Places and Brands

Enhancements

  • Last month we had 4,682,924 places. This month we have 4,774,401 places (net +91,477 places). :chart-with-upwards-trend:

  • We've added 108 new brands :confetti-ball: including:

    • Albertsons Market (albertsonsmarket.com, SG_BRAND_b1cee15316c1cf1a) with 19 places.
    • Abercrombie (abercrombie.com, SG_BRAND_7cced0ecbfbf09fc) with 246 places.
    • DKNY (dkny.com, SG_BRAND_acafddf0ad6cc898f2ff1c5341a4fa38) with 41 places.
    • Loft (loft.com, SG_BRAND_35b7168c01f2e791) with 481 places.
    • Pet Depot (petdepot.net, SG_BRAND_29a84b46d87cd728) with 21 places.
    • and 104 more!!
  • Additional brand parent-child mappings :family: Last month we had 187 parent brands and 409 children brands. This month we have 190 parent brands and 427 children brands. (let us know if we are missing your personal favorite parent-subsidiary corporate structure!). (https://docs.safegraph.com/docs/places-summary-statistics) :chart-with-upwards-trend:

    • For example last month Carl's Jr. (SG_BRAND_a9ecc89d97cef851492e57cd04f9d51b) and Hardee's (SG_BRAND_2b6b8e67fe1433e4f73bbe2c306fb330 ) were not linked in the SafeGraph dataset. This month we now have the parent brand CKE Restaurant Holdings, Inc. (SG_BRAND_c0df4bbabd3f8c58 ) with Carl's Jr. and Hardee's as its children. Siblings reunited! :children-crossing:
SELECT * 
FROM brand_info 
WHERE parent_safegraph_brand_id = 'SG_BRAND_c0df4bbabd3f8c58' 
	OR safegraph_brand_id = 'SG_BRAND_c0df4bbabd3f8c58'
ORDER BY parent_safegraph_brand_id, brand_name

Results:

safegraph_brand_idbrand_nameparent_safegraph_brand_idnaics_codetop_categorysub_category
SG_BRAND_c0df4bbabd3f8c58CKE Restaurant Holdings, Inc.null722513Restaurants and Other Eating PlacesLimited-Service Restaurants
SG_BRAND_a9ecc89d97cef851492e57cd04f9d51bCarl's Jr.SG_BRAND_c0df4bbabd3f8c58722513Restaurants and Other Eating PlacesLimited-Service Restaurants
SG_BRAND_2b6b8e67fe1433e4f73bbe2c306fb330Hardee'sSG_BRAND_c0df4bbabd3f8c58722513Restaurants and Other Eating PlacesLimited-Service Restaurants

Bugs and Known issues

  • Bad SGPID Churn -- One of the hardest problems in maintaining highly accurate and up-to-date monthly releases of SafeGraph Places is the challenge of maintaining consistent, persistent, and unique safegraph_place_id (sgpid). How do we know that a POI in v1 is the same POI in v2? What if they change their phone number? What if their parent shopping mall reconfigures their street addresses? We internally track and estimate our performance in this domain and share these numbers in our release notes for maximum transparency.
    • We dropped 66,025 sgpids (48,276 branded and 17,749 non-branded).
    • We added 157,502 sgpids (91,452 branded and 66,050 non-branded).
    • Some percent of these are true openings and closings; the remainder are bad sgpid churn. We are working on better metrics for distinguishing the two cases.
  • Category Fill Rate We want to have accurate category information for 100% of POI in our dataset; but accurate categories is challenging. We monitor category fill rate with 3 metrics: (1) category fill rate across the entire dataset, (2) category fill rate for branded POI, (3) category fill rate in the brand_info file (brand-level categories). We want all of these numbers to be 100%.
    • (1) All POI category fill rate. Last month 91%. This month 91%.
    • (2) Branded POI category fill rate. Last month 99%. This month 98%.
    • (3) Brand-level category fill rate (brand_info file). Last month 87%. This month 84%.
    • Note: Our Branded POI and Brand-level rates have been slightly decreasing as we have been adding many new brands. We are working to ensure we can continue to add brands at a fast pace without compromising data quality or fill rates.

Geometry

Enhancements

  • Improved and additional cartography and polygons. New or improved polygon geometries :diamond-shape: for over 47,000 POI, including major branded gas stations. Now gas stations include the pumps and lot areas (we've gotten feedback that this is the preferred polygon for visit attribution use cases).

For example, Shell Oil, 915 East Main Street, Livingston, TN 38570 (safegraph_place_id sg:d4279a25b8694690a443f8d71499d811 )

660

Shell Oil, 915 East Main Street, Livingston, TN 38570 (safegraph_place_id sg:d4279a25b8694690a443f8d71499d811)
One of over 47,000 new polygons now available in the SafeGraph Places.

Bugs and Known issues

  • Centroid-Radius Polygons -- A point of pride in SafeGrap Places are the highly accurate polygons that we provide for every POI. However, there are a variety of challenges associated with constructing and maintaining super accurate polygons. Sometimes we do not have sufficient data to represent a POI beyond a point and a circle around that point. We internally track centroid-radius polygons vs precise polygons and are striving for 100% precise polygons. Two months ago we had 90.9% precise polygons, one month ago we had 92.7% precise polygons. This month we have 92.8% precise polygons.