How do I import a CSV with both a billing and a shipping address per row?

Mildport recognizes address blocks and maps them as units: billing columns land together on your billing-address fields and shipping columns on your shipping-address fields — a street from one address is never paired with a city from another. This also works across languages, so 'Rechnungsadresse' columns find billingAddress fields.

What happens when the file only says 'address 1' and 'address 2'?

That is genuinely ambiguous — nothing in the file says which numbered address is billing and which is shipping. Instead of guessing, Mildport asks one clear question per address block ('Which address is address 1?'). One answer maps the whole block, and the answer is remembered, so the next file from the same source maps automatically.

Are 'Address 1' and 'Address 2' treated as two addresses or two lines of one address?

Both conventions exist and Mildport reads the file's structure to tell them apart: standalone 'Address 1'/'Address 2' columns next to a single city and ZIP map to address line 1 and line 2 of one address, while full repeated groups (address1.street, address1.city, address2.street, …) map as separate address blocks.

Does answering an address question once help the next import?

Yes. Answers teach the importer: once you have said that 'address 1' is the billing address, the next file with the same layout maps every address column automatically and asks nothing.

Import rows with multiple addresses (billing, shipping, address 1/2)

Ce contenu n’est pas encore disponible dans votre langue.

Real exports rarely carry just one address. A commerce or CRM file has a billing address and a shipping address; a property-management file has the property’s address, the tenant’s, and the owner’s; plenty of systems simply export address1.street, address1.city, address2.street, address2.city, …. Every one of those columns must land on the right field of the right address.

The failure mode is nasty precisely because it’s quiet: an importer that matches columns one at a time can put the billing street with the shipping city — no error, no warning, just corrupted records that surface weeks later. And an importer that asks the user to untangle it column by column turns two addresses into six questions.

Mildport treats a repeated group as what it is: one thing, decided once.

How multi-address files map

Blocks stay together. Columns that belong to one address map as a unit. A billing street never pairs with a shipping city — if a whole block can’t be placed confidently, it stays unmapped and gets asked about, rather than being scattered.
Meaning over spelling, across languages. Rechnungsadresse and Lieferadresse columns find billingAddress and shippingAddress fields; Dirección de facturación and Adresse de livraison do too. See multilingual CSV import.
Numbered addresses are honest ambiguity. When the file says only address1 / address2 and your schema says billing / shipping, no software can know which is which. Mildport doesn’t guess — and it doesn’t ask six field-level questions either. It asks one question per address block: “Which address is address1?” One click maps the whole block.
Answers are remembered. The question you answered is not asked again. The next file from the same source maps every address column automatically — repeated imports converge to zero questions.
Lines vs blocks, both understood. US-style Address 1 / Address 2 next to a single city and ZIP are the two lines of one address; full repeated groups are separate addresses. The file’s structure decides, not a lucky guess.

The same behavior covers any repeated group, not just addresses — two phone numbers, primary and secondary contacts, property plus owner in one row.

Why this matters for data quality

Address data feeds invoices, deliveries, and legal documents. A silently swapped city is the kind of import bug nobody notices until a parcel — or a dunning letter — goes to the wrong place. Keeping blocks whole, refusing to guess between them, and putting the one real question on the record (with its answer) is what makes an import auditable months later. That record lives in the same question ledger as every other judgment call in the import.

FAQ

How do I import a CSV with billing and shipping addresses in one row? Upload it as-is. Address blocks are recognized and mapped as units — billing columns to billing fields, shipping columns to shipping fields, in any of the six supported languages.

What if the file only says “address 1” and “address 2”? That’s genuinely ambiguous, so Mildport asks one question per address block instead of guessing — and remembers your answer for every future file with that layout.

Are “Address 1”/“Address 2” two addresses or two lines? Both conventions are understood: standalone line columns next to one city/ZIP map as address lines; full repeated groups map as separate addresses.

Does answering once help next time? Yes — one answer maps the whole block and teaches the importer, so the next import of the same layout asks nothing.

See the embeddable importer this runs in, or try a two-address file in the live demo at mildport.com.