The importer is built for the spreadsheet you already have rather than one you tidy up first. It reads your column headers, works out what each one is, and shows you its guesses before anything is written.
It reads CSV only, up to 50MB. Export from Excel, Sheets or Airtable as CSV first.
Running an import
- Open the Locations page and click Batch Import Locations
- Drop in your CSV
- Work through the Smart Mapping step, where the importer shows what it matched each column to and how sure it is
- Read the preview, which renders the first 10 rows as finished location cards
- Start the import
What the mapping step handles
Column headers are matched by meaning rather than exact text, so "Site", "Property" and "Location Name" all reach the name field.
Values get normalised on the way in. "Both" becomes Interior & Exterior. Separate street, city and state columns are combined into one address. Image URLs are pulled out of the wrapped formats that Airtable and similar tools produce.
The Assign mixed tags column with AI toggle deals with the column that holds everything at once. A cell reading "Kitchen, Modern, Pool" describes a room, a style and a feature, and the importer splits it across those three fields rather than dumping the string into one. Mixed Tags AI Analysis covers this in more detail.
Every one of these decisions is visible and overridable in the mapping step. That is where to spend your time, because a column mapped to the wrong field is far more annoying to fix after 300 rows have been written than before.
Large imports take a while
A 300-location file with photos on every row can run for half an hour or more. That is mostly photo fetching rather than the importer being slow, and it is worth knowing before you start one on a laptop you were about to close.
Rows are written in batches, and the batch size drops as the photo count per row rises: ten at a time normally, five for image-heavy files, three for very image-heavy ones. Progress is reported as it goes, so a long import is visibly working rather than apparently hung.
Where it needs your judgment
The importer is good at structure and weaker at meaning. Expect to check:
- Style tags, which are subjective enough that its guesses are worth a read
- Type against Category, since these are separate fields and a spreadsheet that used one column for both will need a decision
- Rates, if your sheet mixed hourly and daily figures in one column, because nothing in the data marks which is which
- Rows with no address, which import cleanly but stay off the map until you give them one
Import 5 or 10 rows first. The preview catches most problems; a small real import catches the rest.