What is POI Data? Points of Interest Data, Sources

POI data, or points of interest data, is a dataset of the places people visit or search for: shops, restaurants, hospitals, schools, parks, and landmarks. Each record pairs a location with attributes such as a name, a category, an address, a phone number, and a website. Retailers, insurers, logistics teams, and map applications use point of interest data to answer where things are, which category each one belongs to, and what sits nearby.

Key takeaways

• A POI record is a point location plus descriptive attributes: name, category, address, contact details, and often a brand and an operating status.
• Open POI datasets now rival commercial ones in size. Overture Maps places holds 81,455,423 places in its September 2026 release.
• POI data comes from business websites, registries, social media, user contributions, and field surveys, then gets conflated to remove duplicates.
• Quality varies by record. Confidence scores, operating status, and stable IDs help filter and track places across releases.
• POIs become useful when joined to other layers: census tracts, drive-time areas, buildings, and hazard maps.

What is POI data?

A point of interest is a term from cartography. The OpenStreetMap wiki describes it as a feature shown with an icon at a single point, such as a post office, a shop, a pub, or a tourist attraction. "Of interest" depends on the task: a postbox matters to someone mailing a letter.

POI data turns those map icons into a table. Overture defines a place as "a concrete, physically identifiable, stationary destination in a publicly observable space." That definition leaves out administrative areas such as cities, movable things such as food carts, waypoints such as bus stops, and private residences.

What a POI record contains

Attribute	Example fields
Identity	Stable ID, source record IDs, release version
Name	Primary name, alternate and translated names
Category	Primary category, category hierarchy, alternate categories
Location	Point geometry (longitude, latitude), sometimes a building polygon
Address	Street, locality, postcode, region, country
Contact	Phone numbers, websites, email, social profiles
Brand	Brand name and a link to a brand registry such as Wikidata
Status	Open, temporarily closed, or permanently closed
Quality	Confidence score, source datasets, last update time

Categories carry most of the analytical value. Overture's taxonomy has roughly 2,300 categories under 13 top-level groups, such as food and drink, health care, shopping, and lodging, plus about 280 simpler "basic" categories for display and search. Foursquare Open Source Places uses more than 1,000 categories.

Commercial schemas add fields for specific workflows. SafeGraph's places schema includes a Placekey identifier, brand IDs, and a polygon geometry for each place.

POI data sources and providers

1. Open datasets. Overture Maps places is free to download as GeoParquet and updates monthly on the AWS Open Data Registry. It is licensed under CDLA Permissive 2.0 and Apache 2.0 and contains no OpenStreetMap data. Foursquare Open Source Places lists more than 100 million POIs in over 200 countries and territories. OpenStreetMap volunteers map POIs as tagged nodes and areas.
2. Commercial POI data providers. Companies such as SafeGraph, Placer.ai, Dataplor, and Mapbox license curated places with verified brands, store polygons, and in some products foot traffic.
3. Map APIs. The Google Places API returns place details, nearby search, and text search results one request at a time. APIs suit applications that look up a few places. Bulk analysis needs a downloadable dataset.

Open data is often built from several providers. Overture's September 2026 places release combines 58,783,121 records from Meta, 10,255,071 from BrightQuery, 6,135,466 from Microsoft, 4,138,835 from Foursquare, 1,809,219 from AllThePlaces, and smaller contributions from DAC, PinMeTo, Krick, and RenderSEO.

How POI data is collected

Placer.ai lists the inputs as websites, social media platforms, government databases, open platforms, and people collecting data in the field. In practice providers combine:

• Business websites and store locators, which list every location of a chain.
• Registries, such as business filings and licensing records.
• Platform data, such as business pages on social networks and review sites.
• User contributions, such as OpenStreetMap edits and Foursquare's Placemaker tools.
• Field surveys, where staff or contractors verify places on the ground.

The hard part is conflation: deciding which records describe the same place. Overture's matcher compares names, addresses, websites, phones, categories, and point locations. Records missing those fields are harder to deduplicate, so they are more likely to persist as duplicates.

POI data quality

POI data changes daily as businesses open, move, and close, so quality is a property of each record. The OpenStreetMap wiki lists five factors that distinguish high-quality POI data: freshness, coverage, consistency, ease of use, and customization.

Overture gives three tools for managing quality:

• Confidence. A score from 0 to 1 for the likelihood that a place exists. Overture documents duplicates, a high junk rate, and low property completeness as known issues, and recommends raising the confidence threshold to trade coverage for precision.
• Operating status. Open, temporarily closed, or permanently closed, where a source reports it.
• Stable IDs. Each place carries a GERS ID that stays the same across releases, so a dataset matched to Overture once can be joined on the ID afterward.

Filtering matters. A Wherobots query of the September 2026 release found 32,370 places in the downtown Chicago area shown in the header image. 3,730 of them had no top-level category, and their average confidence was 0.34. The image keeps places with a confidence of 0.7 or higher and leaves out the services and business group, which holds over half the records.

POI data vs foot traffic data

POI data describes a place. Foot traffic data counts visits to it. Placer.ai builds a polygon around each venue and estimates visits from aggregated, de-identified mobile device location data. POI and foot traffic data are often sold together, and the polygon is the link: to count visits to one coffee shop, you need its boundary, separate from the lot next door.

Common use cases of POI data

• Site selection. Retailers and restaurant chains rank candidate sites by nearby competitors, complementary businesses, and population. The same screen finds data center sites near substations and fiber.
• Trade areas and cannibalization. Drive-time isochrones around each store show which locations compete. Wherobots ran a cannibalization study of two Texas taco chains this way.
• Insurance. Underwriters attach occupancy and nearby hazards to each insured location, then roll exposure into catastrophe modeling. Distance to the nearest fire station is a POI query.
• Logistics and delivery. Carriers geocode stops, find loading points, and measure distances along roads with a spatial join or a nearest-neighbor search.
• Advertising and market research. Analysts count category mixes by neighborhood and link them to demographics from the census.

POI data in Wherobots

Wherobots hosts Overture places as an Apache Iceberg table in the Havasu catalog: wherobots_open_data.overture_maps_foundation.places_place. A count in October 2026 returned 81,455,423 places worldwide, 17,949,232 of them with a US address. Wherobots joined Overture Maps Foundation as a contributing member in July 2024.

Two related datasets save steps. Overture places with isochrones holds precomputed 5, 10, 15, and 20 minute drive-time polygons for 13.3 million US places, described in the isochrones guide. Census tract boundaries and American Community Survey estimates sit in the same catalog. Wherobots has also published a walkthrough of the Foursquare Open Places data.

This query counts food and drink places per 1,000 residents in each Cook County, Illinois census tract:

WITH tracts AS (
  SELECT GEOID, NAMELSAD, geometry
  FROM wherobots_open_data.us_census.tiger_tract
  WHERE STATEFP = '17' AND COUNTYFP = '031'
),
pop AS (
  SELECT geoid, estimate AS population
  FROM wherobots_open_data.us_census.acs_5yr_estimates
  WHERE variable = 'B01003_001' AND geo_level = 'tract' AND vintage = '2020-2024'
),
food AS (
  SELECT geometry
  FROM wherobots_open_data.overture_maps_foundation.places_place
  WHERE taxonomy.hierarchy[0] = 'food_and_drink'
    AND confidence >= 0.7
    AND bbox.xmin BETWEEN -88.27 AND -87.52
    AND bbox.ymin BETWEEN 41.46 AND 42.16
)
SELECT t.GEOID, t.NAMELSAD, p.population,
       COUNT(f.geometry) AS food_places,
       ROUND(COUNT(f.geometry) * 1000.0 / p.population, 1) AS per_1000_residents
FROM tracts t
JOIN pop p ON p.geoid = t.GEOID
LEFT JOIN food f ON ST_Contains(t.geometry, f.geometry)
WHERE p.population >= 500
GROUP BY t.GEOID, t.NAMELSAD, p.population
ORDER BY per_1000_residents DESC

The bounding box filter on bbox prunes the global places table to the Cook County area before the point-in-polygon join. The top tract returns 613 food and drink places for 8,355 residents. Swap the category for health_care or shopping, or swap tracts for drive-time isochrones, and the same pattern answers site selection questions. For the closest place of a type, the nearest-neighbor join pairs each record with its k nearest POIs.

POIs also give AI agents a grounding in real places. The models people use every day were trained on text, documents, databases, and the internet. They can name popular restaurants in a city, and they cannot count the pharmacies inside a 10-minute drive. An agent connected to the Wherobots MCP server can answer that with a spatial query.

Related terms

• Spatial join: matching POIs to tracts, isochrones, and parcels by location
• Data center site selection: screening sites by power, water, and nearby infrastructure
• Parcel data: the land records behind each address
• Building footprint: the outlines of the buildings that hold each place
• Raster vs vector data: POIs are vector points
• Physical AI: AI that works with real places
• H3: a hexagonal grid for counting POIs per cell
• Geospatial analysis: density, proximity, and clustering methods applied to POIs

Read more from Wherobots

• Location intelligence with isochrones: a complete guide
• Comparing taco chains: a consumer retail cannibalization study with isochrones
• Tutorial: how to use GERS IDs in Wherobots for geospatial data integration
• Exploring the Foursquare Open Places data with Wherobots

Query Overture places, census tracts, and isochrones with a Wherobots free trial at cloud.wherobots.com.