A point of interest (POI) is a defined location that matters for analysis — a shop, restaurant, transport hub, venue or landmark. In a location dataset, POIs are the reference geography that raw coordinates are measured against: without them, a position is just a number pair.
Points, polygons and why the difference matters
A POI is often stored as a single coordinate, but a single point cannot say whether someone was inside a building or walking past it. Serious visitation analysis uses polygons — boundaries drawn around the actual premises. Polygon quality is the single largest driver of accuracy in visit attribution: a boundary that spills into a neighbouring unit, a pavement or a shared car park will record visits that never took place.
What a POI record usually contains
Beyond geometry, a usable POI dataset carries a name, a category or classification, an address, and often opening hours, brand or chain membership, and a stable identifier so the same place can be tracked over time.
Keeping them current
POI data decays quickly. Businesses close, rebrand, move and change footprint, and a dataset that is not actively maintained will silently misattribute activity to places that no longer exist. Recency of update is a fair question to ask any POI supplier.
Leave a Reply