Mirakl · Moving in
Move from Mirakl to a Parody
Export your records from Mirakl, import them into a Mirakl Parody with CSV files, check them, then switch over. You can move back the same way: Move back to Mirakl.
Before you start#
Mirakl is run for each marketplace by its operator, and bulk data comes out through Mirakl's APIs rather than one-click exports. You'll pull shops, the catalog, offers, and orders with the operator APIs, turn the JSON into CSV where needed, and import the files into your Parody.
Who can do this
You need an operator API key for your Mirakl instance. Your Mirakl contact can also arrange a full data extract; check what your contract includes before you start.
What comes over
| In Mirakl | In your Parody |
|---|---|
| Shops (sellers) | Shops |
| Catalog categories (hierarchies) | Categories |
| Products | Products |
| Offers | Offers |
| Orders | Orders |
What doesn't
- Order lines, shipments, refunds, and messages. They can be added from the order data with more preparation; this guide covers order headers.
- Payments, invoices, and accounting documents. Import them as records for reference; no money moves.
- KYC and shop documents as files, and seller logins.
- Storefront and seller integrations. Hosted Parodies have no APIs.
- Anything the Mirakl Parody has no field for. Check the column tables below; columns without a match are left out.
1. Set up your Parody#
- Create a Mirakl Parody with at least these features, so every record type has a home: Sellers, Category hierarchy, Products, Offers, Orders. See what's in the Mirakl Parody.
- New Parodies come with sample records. Open Import & export in your Parody, and under Start empty, delete them.
2. Export from Mirakl#
Call the APIs below against your instance. Each is documented on developer.mirakl.com. For JSON responses, flatten them to CSV with a spreadsheet's JSON import, a script, or a tool such as jq.
OF51 is going away
The older offers export, OF51, is deprecated. Use the asynchronous OF52, OF53, and OF54 instead.
What to export:
- Shops (sellers): the shops list API,
S20, which returns shops as JSON, page by page. - Catalog categories (hierarchies): the back office's catalog settings or the hierarchy APIs. You need each category's code, label, and parent code.
- Products: the product export API,
CM51. - Offers: the asynchronous offer export: start it with
OF52, poll withOF53, and download the files withOF54. Ask for CSV. - Orders: the asynchronous order export:
OR13to start,OR14to check its status,OR15to download.OR11lists orders page by page for smaller volumes.
3. Prepare the files#
You need one CSV per record type. Open each export in a spreadsheet app, make the changes below, and save it as CSV (UTF-8). Keep the header row.
Mirakl's IDs (shop IDs, product SKUs, offer IDs, order IDs) are what link records together, so map each file's own ID column to ID. Offers then find their shop and product, and orders their shop.
Keep the ID column
Map Mirakl's own ID to ID. Your Parody remembers it, so later files can link to the record by that ID, and importing a fresh export later updates records instead of duplicating them.
Shops
| Mirakl column | Parody field |
|---|---|
shop_id | ID (match and link rows) |
shop_name | Name |
contact email | |
shop_state | Status |
country | Country |
Categories
| Mirakl column | Parody field |
|---|---|
Category code | ID (match and link rows) |
Label | Name |
Parent code | Parent category |
Sort parents above their children, so each row's parent already exists when it's read.
Products
| Mirakl column | Parody field |
|---|---|
Product SKU | ID (match and link rows) |
Title | Title |
Brand | Brand |
EAN | Ean |
Category code | Category |
Description | Description |
Offers
| Mirakl column | Parody field |
|---|---|
offer-id | ID (match and link rows) |
shop-id | Shop |
product-sku | Product |
price | Price |
discount-price | Discount price |
quantity | Quantity |
state-code | Condition |
leadtime-to-ship | Lead time days |
Column names vary with your Mirakl version and export options. Match them in the import screen if they differ.
Orders
| Mirakl column | Parody field |
|---|---|
order_id | ID (match and link rows) |
shop_id | Shop |
customer name | Customer name |
total_price | Total |
order_state | Status |
created_date | Order date |
shipping_deadline | Ship by date |
4. Import, in this order#
Import parents before the records that link to them, so links resolve:
- Shops
- Categories
- Products
- Offers
- Orders
For each file:
- In your Parody, open Import & export.
- Under Import, choose the record type and the CSV file.
- Check the column list. Columns named like the tables above are matched for you; pick a field for any that aren't, or leave them out.
- Select Import. If any row has a problem, nothing is written and you get a list of the problems by line number. Fix the file and import again, or tick Skip rows with problems.
Files can hold up to 5,000 rows and 4 MB. Split bigger exports and import the parts one after another. The full rules are in Import & export.
5. Check the result#
- Compare the counts on your Parody's overview with the row counts of your files.
- Open a few records of each type and check their links (the project on a task, the account on a contact).
- Open the boards and reports and check that records sit in the right columns.
6. Switch over#
- Pick a cutover time and ask your team to stop editing in Mirakl.
- Export again and import the new files. Rows with an ID you've imported before update those records; new rows are added.
- Share the Parody's address with your team, and keep Mirakl read-only until you're sure nothing is missing.
Troubleshooting#
| Message | What to do |
|---|---|
| No … named or with id "…" | The record it links to isn't in the Parody yet. Import that record type first, or check the name is spelled exactly the same. |
| "…" matches 2 … records | Several records share that name. Also map the column that links to their parent (the project or board), or use the ID instead of the name. |
| … is not a date we can read | Use YYYY-MM-DD. Day-first dates like 21/11/2026 aren't read, because 03/04/2026 would be ambiguous. |
| … is not a number | Remove words and units other than currency symbols, commas, and %. |
| The file has more than … rows | Split it into files of up to 5,000 rows. |
Sources
Checked against these pages on 2026-11-21. Vendors change their export and import tools often; if a step doesn't match what you see, their page wins.