Point of interest data
POI Data Services
Business and point-of-interest data with geographic coordinates, consistent categories and documented coverage.
- Restaurants · 34
- Cafés · 22
- Retail · 26
Overview
A POI dataset is only as good as its category scheme
Points of interest are simple on the surface: a name, a coordinate, a category. The difficulty is consistency. One source calls a venue a “cafe”, another “coffee shop”, a third “restaurant”. Count them naively and your market size moves by a third depending on which label you trusted.
We standardise categories into a scheme you can defend, keep the original source label in its own column, and document exactly how the mapping was made. Nothing is silently reclassified.
The same discipline applies to duplicates. The same venue appearing in two sources is resolved with normalised name matching plus a distance threshold, and the merge rules are written down rather than buried in a script.
Real output
A real extract, mapped and measured
8,628 POI records across Manhattan, taken from a single OpenStreetMap extract and put through the same steps a delivered dataset goes through. The point is not the map; it is that every claim below can be checked against the data.
Every record, where it actually is
One dot per record, coloured by category, with the street network underneath for reference. Plotting the whole extract rather than a sample is the first honest check on a POI dataset: gaps, duplicates and misplaced coordinates all become visible at this zoom.
The same data at street level
Zoomed to Midtown, where 3,125 of those records sit. Coordinates are precise enough that records land on the correct side of the street, which is what makes catchment and frontage analysis possible later.
How complete each field really is
Measured across all 8,628 records rather than estimated. Name, coordinates and category are effectively universal; phone, website and opening hours sit near half. We report this per project so you know what you can build on before you build on it.
Data © OpenStreetMap contributors, available under the Open Database Licence. Used here as a public dataset we can show openly; client work runs on the sources agreed for that project.
Scope
What a POI record contains
A standard record. Additional attributes can be added where the source publishes them.
POI identifier
A stable key that survives refreshes, so you can track a venue over time.
Name
Published name, plus a normalised form used for matching.
Standardised category
Mapped to an agreed scheme, with the original source label retained alongside it.
Coordinates
Latitude and longitude in WGS 84, validated against the address.
Address components
Street, city, region, postcode and country, parsed rather than left as one string.
Contact details
Business phone and website where published by the business.
Opening hours
Weekly schedule in a structured form, including seasonal and closed-day handling.
Popularity signals
Rating and review counts where published, useful as a demand proxy.
Status
Open, temporarily closed or permanently closed, so counts stay honest.
Brand and chain flag
Whether a venue belongs to a chain, and which one, for competitive analysis.
Administrative joins
Census tract, postcode area or custom zone identifiers, pre-joined for aggregation.
Collection date
When the record was captured, so freshness is never in question.
Output formats
Delivered the way your stack expects
- CSV
- One row per POI, with coordinates as decimal degrees.
- Excel
- Workbook with data, category mapping and coverage sheets.
- JSON
- Nested records preserving hours, categories and attribute arrays.
- GeoJSON
- Point features in EPSG:4326 for web maps and GIS software.
- Shapefile
- Point shapefile with projection metadata for ArcGIS workflows.
Use cases
What clients use it for
- Catchment profilingDescribe what surrounds a location: transit, schools, offices, food and retail.
- Delivery and logistics planningBuild accurate destination sets for routing, zoning and coverage models.
- Heatmap inputsFeed density and distribution analysis with a clean, deduplicated point layer.
- Market entry researchCompare category structure between cities before choosing where to launch.
Sample dataset
What you actually receive
Sample project data. The source label stays in the dataset so any category decision can be audited or reversed.
| POI ID | Name | Standard category | Source label | Latitude | Longitude | Status |
|---|---|---|---|---|---|---|
| POI-8841021 | Kirkwood Coffee | Cafe | Coffee shop | 33.75121 | -84.31688 | Open |
| POI-8841022 | Grant Park Grocers | Grocery | Supermarket | 33.73894 | -84.36002 | Open |
| POI-8841023 | Edgewood Pizza Co. | Restaurant · Pizza | Pizza restaurant | 33.75630 | -84.34199 | Open |
| POI-8841024 | Old Fourth Ward Gym | Fitness | Gym | 33.76412 | -84.36871 | Temporarily closed |
| POI-8841025 | Inman Park Bakery | Bakery | Bakery | 33.76187 | -84.35364 | Open |
| POI-8841026 | Reynoldstown Bar | Bar | Cocktail bar | 33.75208 | -84.34617 | Open |
How it works
Five steps, every project
- Step 01
Tell us what data you need
Which categories, which geography, and what the counts will be used for.
- Step 02
We define the data scope
We agree the category scheme, the field list and how chains and duplicates are treated.
- Step 03
We collect and process the data
Sources are collected, addresses parsed, coordinates validated and categories mapped.
- Step 04
We validate the dataset
Duplicate rate, category coverage and spatial distribution are checked against expectations.
- Step 05
We deliver the final result
POI layer plus category mapping table, coverage summary and collection dates.
Quoted on volume, the number of categories and cities, and the refresh cadence if the dataset is a recurring one. Tell us the coverage you need and you get a fixed price against a written scope.
FAQ
Questions we get asked
What is POI data?
POI stands for point of interest: a specific place recorded as a coordinate with attributes, such as a restaurant, pharmacy, ATM, school or transit stop. In practice a POI dataset is a table of places with a name, a category, a location and whatever attributes the source publishes.
How many POIs can you collect for a city?
It depends on the categories. A single category in a mid-sized metro is typically a few thousand records; a broad multi-category sweep of a large metro runs into the tens of thousands. We give you a count estimate from a pilot area before the full run.
How do you handle duplicate venues?
With a documented rule set: names are normalised, then candidate matches within a distance threshold are compared on address and category. Merges keep the most complete record and retain the alternate source identifiers so nothing is lost.
Can you use our own category scheme?
Yes. If you supply a taxonomy we map to it and deliver the mapping table so you can audit every decision. If you do not have one, we propose a scheme based on how you intend to aggregate the data.
How fresh is the data?
Every record carries a collection date. For a one-off delivery, the data is as fresh as the run. For monitored categories, weekly or monthly refreshes keep status changes such as closures current, with a change log per run.
Can you add attributes to POIs we already have?
Yes. Send us your list with whatever identifiers you use and we return it enriched with coordinates, categories, catchment attributes or administrative joins, keeping your keys intact.
More services
Often combined with
Street View Data
Large-scale street view imagery collection, processing and structured dataset delivery.
Learn MoreSemantic Segmentation
Pixel-level image analysis that turns imagery into counted, measured classes.
Learn MoreGoogle Maps Data
Structured business listings with coordinates, categories, ratings and opening hours.
Learn MoreKeep exploring
Where to go next
Related services
- Google Maps DataStructured business listings with coordinates, categories, ratings and opening hours.
- Heatmap AnalysisDensity, distribution and hotspot analysis that stands up to statistical scrutiny.
- Location IntelligenceLocation-based market, catchment and competitor analysis that ends in a recommendation.
- Data CollectionCustom data collection and structured datasets built around one specific question.
Related reading
Next step
Have a specific data requirement?
Tell us the geography, the fields and the cadence you need for poi data. You get a scoped plan, a sample and a fixed price before any work starts.