Import & export
Anything you keep in a spreadsheet can come in as a CSV: items, opening stock, locations, products and their variants, kits, suppliers and customers, catalog prices, purchase orders and sales orders. Much of what is in Manifold can go back out the same way.
Moving from Cin7? You do not need CSV files: paste a Cin7 key and Manifold reads the account across. See Moving from Cin7. New to Manifold? Start with Getting started.
Open Import in the sidebar. The screen has two tabs, Import and Export. Owners, admins and people with the Data migration role can use both. Exporting the cost of goods sold ledger or stock valuation also needs the Finance role.
Importing a file
- Select Choose a CSV file. Leave Import as on Detect from the file's columns, or pick what the file holds.
- Check what it was Recognised as (it reads Importing as if you picked the type yourself). Each choice in that list says how many of your columns it recognises. Pick another if the guess is wrong.
- Check the column mapping above the preview of the first 50 rows. Each column is matched to a field from its header (SKU, Name, Qty and so on). Change any that are wrong, or set them to Skip this column. Required fields are marked *. Untick First row is a header if your file has none.
- Fix what the screen flags. Missing required fields are listed above the buttons. Cells that would be rejected are shown in red with the reason.
- Choose Start import.
Progress shows rows Applied, Failed, the total and the rate, with the time remaining. You can close the tab: the import keeps running.
When it finishes, Rejected rows lists each failed row with the field and the reason. Download as CSV gives a list of each failed row's number, field and reason, not the rows themselves. Open it beside your source file, correct those rows in the source, and import only the corrected rows as a new file. A failed row does not stop the others. If the import as a whole stops, the screen says why, and rows already applied stay applied.
If you import the same layout often, type a name and choose Save for reuse. Next time, pick it under Reuse a saved mapping.
What can come in
- Locations
- The places stock is held, one row each.
- Inventory items (SKU master)
- One row per SKU, with an opening quantity. A serial- or batch-tracked SKU must come in at 0 and be received against a purchase order, so every unit is identified.
- Stock adjustments
- A quantity to add or remove (negative to remove) for a SKU that already exists, with a reason.
- Products and variants
- One row per variant, grouped into a product by a key column. The product's options are read from the option columns. Each variant's SKU must already be an inventory item.
- Composite variants (kits)
- A kit's recipe, one row per component.
- Catalog prices
- A price per SKU per catalog. A blank price removes it. The catalog must already exist.
- Companies and People
- Suppliers, business customers and contacts.
- Purchase orders
- One row per line, grouped by your PO number. They come in as drafts, for you to authorise.
- Goods receipts
- Deliveries against an authorised purchase order.
- Supplier payments
- Money paid to a supplier against what its purchase orders owe.
- Sales orders
- One row per line, grouped by reference. Stock is allocated as each order comes in.
- Shipment picks, Shipment despatches, Refunds
- Steps on parcels and orders that already exist in Manifold.
- Historical orders
- Orders from before Manifold. They hold no stock and ship nothing; they are your sales history.
- Payouts
- A payment provider's payout report, matched to the payments and refunds it settles.
Not importable: catalogs themselves, authorising a purchase order, packing, invoices, payments against sales orders, transfers and users. Do those in the admin.
A catalog pulled from Shopify or a migration from another system comes in through the same import, so it is checked and reported the same way as a file.
The order to load in
Each import type needs what the ones before it create:
- Locations.
- Inventory items. Stock adjustments can follow any time after.
- Products and variants, then kits.
- Create your catalogs in the admin (Catalogs).
- Catalog prices.
- Companies and people. Historical orders can come in any time after.
- Purchase orders. Authorise them, then import goods receipts and supplier payments.
- Sales orders, then picks and despatches, then refunds, then payouts.
Writing the file
- Money is a whole number of hundredths, for every currency:
1250for 12.50, and150000for ¥1,500.12.50is rejected. - SKUs are stored in upper case, whatever case the file uses.
- Keep a document's rows together. All lines of one purchase order, sales order or refund should sit next to each other.
- Rows are applied 500 at a time. A document whose rows cross a 500-row boundary loses the lines after it. In large files, keep each document's rows well inside one block of 500, or split the file.
- List cells, such as tags, are separated by
;or,. - An import must finish within about 15 minutes. Rows it reaches after that are refused and listed under Rejected rows. Watch the rate and time remaining as it runs; if the time remaining is near 15 minutes, split the file into several imports.
Importing the same file again
Each row becomes the same change you would make by hand, recorded under your name in the activity log. Most files are safe to import twice: the second time, rows already applied change nothing. They either count as applied again or show under Rejected rows as already existing.
Three are not. Importing these again applies them again:
- Stock adjustments, always. Never import an adjustment file twice.
- Goods receipts without a receipt id column are received again, up to whatever is still due on the purchase order.
- Companies without a reference column. Each one is registered twice.
Fill in the reference or receipt id column wherever an import type offers one.
While your organization is read-only and kept in step with the system you imported from, products, stock, orders and the other imported areas can't be changed here, so rows for them are refused. See Going live.
Exporting
- Open the Export tab and choose what you are exporting.
- The preview shows the file, 25 rows to a page. Use each column's filter (Filter by condition) to narrow the rows, and the Columns menu to choose columns.
- Choose Export CSV. The file downloads with exactly the rows and columns the preview shows.
You can export products and variants, catalog prices, inventory items, locations, purchase orders, sales orders, the cost of goods sold ledger and stock valuation.
Exported columns use the importer's names, so a file can go out, be edited and come back without mapping. Catalog prices are the usual round trip. A blank cell means none, not zero.
If your subscription lapses, you can still export your data.
For developers
- Flow: bulk import from a spreadsheet: upload, map, run and read the failures.
- Import / Export API: every target and its fields, and rows sent as JSON.
- Idempotency and retries: why reference columns make a re-import safe.