How to enrich addresses with business data when the list has no names
Turn a bare column of street addresses into businesses with phone, website, rating and type: place each address, find what sits there, keep the closest match.
Part of the guide Location intelligence automation
To enrich addresses with business data when all you have is the address, first give each address a position on the map, then look for the businesses within a few metres of that position, keep the closest one (or all of them, for a building with several tenants), and copy across its name, type, phone, website and rating. Keep the distance between the address and the business you matched, because that one number tells you how far to trust each row.
Lists like this turn up more often than you would think: a landlord's schedule of commercial units, a council's list of licensed premises, delivery drop points from a logistics system, addresses typed by field staff. This guide covers the steps in order, a made-up example, the ways matching goes wrong and how to have an AI employee do it again whenever the list grows. The workflow automation guide covers the general idea of chaining steps like these.
Why addresses alone are harder than names
Enriching a list of business names is a lookup: search for the name near its town. An address is the other way round. You know where, not who. Several businesses can share one address (a mall, an office tower, a row of shop units with one number), a business can have moved out since the list was written, and the address may be written in a way no map recognises. The method below works with that uncertainty instead of hiding it.
The steps
- Clean the address column. One address per row, with the street, number, city and, if you have it, the postcode in separate columns or at least in a consistent order.
- Give each row a position. Upload the list as a CSV on the My data page. Rows without latitude and longitude are placed from their address and city columns.
- Search a small circle around each position. In a flow, start with My data on that table, then Nearby places set to Each incoming place, a distance of 100 m, passing on the places found. Leave Types of place empty to find every kind of place, or pick the kinds you expect.
- Keep the closest match. Every place found carries near (which address it was found around) and distance_m. Sort and limit by distance, then Remove duplicates on the near field keeps the nearest business for each address. Skip this step if you want every tenant.
- Mark how sure each match is. A Calculated field can label each row: under 25 m a likely match, 25 to 100 m worth a look.
- Send the result where it is used. A sheet, your CRM, or back into My data as a new table.
Each match brings what xMap's places database knows about the place: name, type and its wider group, brand and whether it is a chain, rating and number of reviews, phone and website when known, opening date, and whether it is still open.
A worked example
This example is made up. A pest control company in Cape Town has 280 service addresses from an old billing system, with no business names: just street, number and suburb. Before a sales push, the manager wants to know what business sits at each address today and how to reach it.
- She exports the addresses as a CSV and uploads it on the My data page; the rows are placed from their address and suburb.
- A flow reads the table, runs Nearby places within 100 m of each address with no type picked, sorts by distance and keeps the nearest place per address.
- A Calculated field labels each row Likely (under 25 m) or Check (25 m and over).
- The result goes to a new sheet, sorted so the Check rows are at the top.
In this made-up run, most addresses come back with a restaurant, shop or clinic a few metres away; a handful in a shopping centre come back with several tenants at the same spot, and some show a place marked closed or nothing at all. Those last two groups are exactly the rows worth a phone call or a drive past: the customer may have moved.
Where address matching goes wrong
- Shared buildings. One address, many businesses. The nearest is not always the one you billed. If you know the kind of business you served, pick those types in Nearby places.
- Positions placed at a street's middle. A vague address can be placed roughly, so the nearest business may be next door. A larger distance in the result is the warning sign.
- Closed places. Nearby places leaves out places marked closed by default. Switch that off for this job if you want to know which addresses now hold a closed business.
- Missing fields. Not every place has a phone or website. Count the gaps before promising a full call list.
- Searching too wide. A 500 m circle in a city centre finds hundreds of places and the nearest may be across the road. Start small.
Doing it again whenever the list grows
Once the flow works, make it run on its own. If the addresses arrive in a shared sheet, start the flow with the New rows trigger so every address typed in is enriched within minutes. The template Enrich a lead sheet and push it to HubSpot shows that pattern with names; Who runs the businesses around my stores starts from your own table of addresses and adds who to talk to at each business nearby. The AI employee leaves a note in its Team inbox thread after each run with how many addresses were matched and how many need a look.
Questions
Do I need latitude and longitude in my list?
No. Upload the list on the My data page and rows without a position are placed from their address and city columns.
What distance should I search around each address?
Start with 100 m. In dense city centres go lower; for business parks with large plots you may need a little more.
What if several businesses share the address?
Keep them all, or keep the nearest. Each place found says which address it was found around and how far from it, so you can decide row by row.
Will it tell me if a business has closed?
Places marked closed are left out by default. Switch off Only places that are open to see them, with their status.
Can this run automatically when new addresses are added?
Yes. Start the flow with New rows on a Google Sheets sheet or a My data table, and each new address is enriched shortly after it is added.
Templates that do this
Read next
More on business and places data · Every article · Templates