How to move products, variants, customers and orders into Shopware 6 via CSV, and which mistakes to avoid along the way.
The Shopware CSV import is a built-in module in Shopware 6 that lets you bring products, customers, categories and other data into your store from a CSV file. You find it in the Administration under Settings, Automation, Import/Export. Every file runs through a profile that defines which column goes into which Shopware field. The file must be UTF-8 encoded, use semicolons as separators and dots as decimal marks.
For individual data areas CSV is the fastest route, for a complete platform move it usually is not. Products, prices, stock levels and customer lists work well via CSV. Order history, reviews, SEO URLs and documents move more cleanly with the Migration Assistant when you come from Shopware 5 or another Shopware 6 installation. Ongoing syncs with an ERP or inventory system belong in an integration.
Most import errors happen before the upload. Excel changes special characters, separators and leading zeros, variants need a fixed order, and customers receive registration emails unless you prepare for it. If you would rather not handle the import yourself: in the solution25 store subscription we create up to 200 or 500 products depending on the plan, and the Enterprise plan includes data migration.
This article is for store owners, ecommerce managers and developers who need to bring data into a Shopware 6 store: when launching a new store, during a relaunch or for the regular upkeep of large catalogs. You will learn which method fits which case, how to prepare profiles and files, and how to import products, variants, customers and orders without losing data.
The Import/Export module is part of the Shopware 6 standard feature set. You do not need an extension for it. In the Administration, open Settings, Automation, Import/Export. There are three tabs: Import, Export and Profiles.
Shopware ships default profiles for products, categories, properties, newsletter recipients and the variant configuration, among others. The documentation also contains field references for customers, orders, media, prices, cross-selling and promotion codes. You can create your own profiles for each of these object types.
When you start an import, Shopware validates every record. Valid rows are imported, invalid rows end up in a separate CSV file with an additional _error column. You download that file, fix it and import it again. The log shows the imports of the last 30 days, including the source file, the profile and the error file.
An important button is the dry run. It runs the complete import without writing any data. This shows you in advance whether rows will fail, so you avoid a partially imported data set.
An import can only add or overwrite data, not remove it. An example from the documentation: once a sales channel assignment is set, you cannot delete it again via import. To clean up assignments, use the Administration or the API. Aborting an import does not undo anything either: records written up to that point stay in the system.
Before you build the first file, check whether CSV is the right tool at all. Shopware offers three ways to bring data into the store, and each fits a different case.
| Method | Best for | Limits |
|---|---|---|
| CSV Import/Export | Initial catalog setup, bulk changes to prices and stock, customer lists from a legacy system | No deletions, variants only in several steps, no standard guide for orders |
| Migration Assistant | Moving from Shopware 5 or another Shopware 6, Magento according to the Shopware documentation | Payment methods, flows, dynamic product groups and extension data must be set up again |
| API or integration | Ongoing sync with inventory, ERP or PIM | Development effort, operation and monitoring of the integration |
CSV is strong when data arrives once or at larger intervals and the structure is manageable. Typical cases: a retailer launches with 800 items from a supplier list. A store changes the prices of 2,000 products before Black Friday. A wholesaler moves its customer list from an old B2B portal. In all these cases a well-built CSV is imported within minutes.
If you come from Shopware 5, use the official Migration Assistant. According to Shopware it moves media, categories, products, properties, manufacturers, reviews, cross-selling, customers, orders including documents, SEO URLs, newsletter recipients, wishlists and promotions. Rebuilding all that via CSV takes many times longer and loses relationships between the data. How this move works is described in the article Migrating from Shopware 5 to Shopware 6.
As soon as data has to flow daily or hourly, for example stock levels from the warehouse or orders back to accounting, manual CSV files are the wrong approach. Then you need a connection to your inventory system. Which systems exist and how the connection works is covered in the article Shopware inventory management. An intermediate solution is automated CSV imports via the command line, more on that below.
Shopware expects a real CSV file with fixed rules. The documentation names three: UTF-8 as character encoding, semicolon as field separator, quotation marks around text. For numbers, add the dot as decimal separator, so 5.00 and not 5,00. US teams are used to the dot already, but the semicolon trips up many exports that default to commas.
Most retailers build their lists in Excel. When saving as CSV, three typical things happen: special characters are encoded incorrectly, semicolons turn into commas or vice versa, and leading zeros disappear. Item number 00451 becomes 451, and a 13 digit EAN or a 12 digit UPC turns into a number in scientific notation. Shopware therefore recommends a program that writes native CSV files and names OpenOffice Calc as an example. LibreOffice Calc or a text editor with encoding selection work just as well.
The most reliable way to the right structure: create two or three test products by hand, export them with the profile you will later use for the import, and use the exported file as a template. Column names, order and format will then be correct. The Shopware forum recommends exactly this approach for the error message about unsupported file formats. Several users there report that the same upload failed in Firefox and worked in Chrome.
A profile is the translation between your file and the Shopware database. You can duplicate a default profile or create a new one. Important: select the system default language when you do, otherwise the profile will not work according to the documentation.
These two switches are a simple safeguard. For a pure price update, turn off record creation. A typo in an item number then does not create a new, empty product but ends up as an error in the error file.
In the mapping you assign a column name to each database field. For each row you set whether the field is required, which default value it gets and at which position it sits. If you upload an existing CSV, Shopware suggests the fields from the header row. The system adds the required fields itself:
| Object type | Required fields according to the documentation |
|---|---|
| Product | id, taxId, productNumber, stock, name |
| Customer | id, defaultBillingAddressId, defaultShippingAddressId, customerNumber, firstName, lastName, email |
| Category | id, type, name |
Internally, Shopware works with long UUIDs. No legacy system knows these IDs. In the advanced profile settings you can therefore define a second unique identifier, for products typically the item number (productNumber). Shopware then finds the right product during an update via the number. Make sure this identifier really is unique. Duplicate values cause records to be skipped.
The product import is the most common case and the one with the most hidden details. With the default product profile you get far if you watch the following points.
Price fields follow the pattern price.DEFAULT.net and price.DEFAULT.gross. DEFAULT stands for the default currency. Provide both values or decide beforehand which value leads. A common mistake is a gross price that is calculated back with the wrong tax rate, leaving odd cent amounts in the store.
The product file has the columns tax_id, tax_rate and tax_name. For tax_id, enter the UUID of an existing tax rate. If the field stays empty, Shopware creates a new tax rate according to the documentation. That sounds harmless but quickly leads to several duplicate rates with the same percentage that nobody can tell apart later. Get the UUIDs beforehand via an export or from the Administration.
The media field takes the URLs of the product images, the cover image has its own field. Shopware downloads the images from these addresses. The images must therefore be publicly reachable. Links from cloud shares with an expiry date or from a supplier portal behind a login fail. For large catalogs a dedicated image folder on a server pays off, which you empty again after the import.
You assign products via category IDs. Create the category structure before the products, either by hand or with the default category profile. There, parent_id defines the parent category. You see a category ID in the Administration address bar when you open the category.
Text in other languages runs through fields following the pattern translations.DEFAULT, where DEFAULT is replaced by the locale code, for example de-DE or en-GB. For a multilingual store, plan one column per language for name, description and meta data. Shopware recommends a database backup before importing translations.
Variants are where most CSV imports fail. The reason: a variant depends on a main product and on properties that both have to exist first. The Shopware documentation therefore describes four steps.
When importing variants, parent_id, product_number, active, stock and optionIds are required. Separate multiple option IDs with a pipe character, so id-red|id-xl.
Generating is faster when every combination should exist and the variants differ only in their options. Importing is better when you need to bring over your own item numbers, EAN, stock or prices per variant. In practice we often combine both: generate the variants, then add numbers, stock and prices with an update import. This is our project experience, not a Shopware requirement.
A hypothetical fashion retailer has 300 shirts in 5 sizes and 4 colors each. That makes 300 main products, 2 property groups with 9 options, 2,700 rows of variant configuration (300 times 9) and 6,000 variants (300 times 20). Packing all of that into a single file will fail. In four cleanly separated files it is a manageable import.
A customer import is about more than names and addresses. Three points decide whether the move goes unnoticed by your customers.
Shopware sends a registration email when a customer is created. With 5,000 imported customers that would be 5,000 emails to people who never registered again. The documentation therefore advises disabling mail delivery under Settings, System, Mailer before the import and turning it back on afterward.
Customers should be able to log in with their old password after the move. Shopware provides the fields legacyPassword and legacyEncoder for this. They hold the encrypted hash from the legacy system and the method used to create it. When the customer logs in, Shopware checks the old password and saves it again in its own format. Plain text passwords have no place in an import file. If Shopware does not support the legacy method, the only option is a password reset email after go-live.
Every customer needs a default billing and a default shipping address. For B2B customers, customer group, company name and tax ID are added. Create the customer groups beforehand and check whether countries and salutations in the legacy system have the same values as in Shopware.
If you sell to customers in the US, you usually do not maintain a fixed tax rate per product in Shopware the way European stores do. Sales tax depends on the state and often on local jurisdictions, so it is calculated at checkout through an integration with a tax service. In the import file you assign each product a tax class and let the service handle the rates. If your customer list includes customers in the EU, the GDPR applies to their data. Most Shopware partners are based in Europe, so plan imports and go-live around the time difference: the overlap with Central European working hours is usually your morning on the East Coast.
Regardless of the country: only bring over customers you still need. A move is a good opportunity to remove inactive accounts according to your retention periods. Import files with customer data do not belong in email attachments or open cloud folders and should be deleted after the import. More on this in the article Data protection in online stores.
For orders, the Shopware documentation contains a field reference but no step-by-step guide like the one for products or variants. As minimum fields it names id, salesChannelId, orderDateTime and stateId. On top of that come line items, addresses, payment and shipping status, each attached to their own tables.
Our recommendation from projects: do not move order history via CSV if there is another way.
If you are planning a move from another platform, take a look at the article Switching from Shopware to Shopify. It shows from the opposite direction which data typically gets lost when changing platforms.
When you fill a new store, the order decides the effort. Each step needs IDs from the previous one. This sequence has proven itself:
The import itself is rarely the time problem. The work is in preparing the data. As a rough guide from our projects, with no guarantee for your case:
| Starting point | Data preparation effort |
|---|---|
| Up to 500 simple products, clean supplier list | 1 to 2 working days |
| 2,000 products with variants and images from several sources | 1 to 2 weeks |
| Complete move with customers, variants and translations | Part of a migration project, planned in weeks |
The biggest time sink is almost always the quality of the source data: missing item numbers, duplicate EAN, sizes sometimes as M, sometimes as Medium. Plan a cleanup before you build the first file. A test instance where you can safely repeat imports saves time in every project. More on versions and updates of the test environment in the article Shopware updates and versions.
For recurring imports you do not have to open the Administration every time. Shopware has a console command import:entity, which according to the command reference imports entities from a CSV file. The exact arguments are shown by bin/console import:entity --help on your installed version. This lets you, for example, load a stock list from your supplier every night, controlled by a cron job on the server.
Keep two points in mind:
messenger:consume, follow-up work such as indexing stays pending. Properly configured hosting takes care of this, more in the article Shopware hosting.import-export:delete-expired deletes expired import and export files. Especially with customer data, these should not stay on the server permanently.When an automated CSV import serves more than one data source or has to run in both directions, it is usually the forerunner of a real integration. From that point on, a connection to the inventory system or ERP pays off.
These are the mistakes we see most often in projects. Most can be avoided with ten minutes of preparation.
An import writes immediately. A wrongly mapped price field changes thousands of prices at once, and aborting does not undo anything. Test every new file with the dry run first and on a test environment. A database backup belongs before larger updates.
If you leave the ID column empty during an update and have not set a second identifier such as the item number, you get new products instead of updated ones. In the end there are duplicates in the store. Turn off record creation in the profile for pure updates.
Shopware itself points out that the message after an import does not mean everything worked. Check in the log how many records failed and download the error file.
An import adds and overwrites, it does not delete. If a product should disappear from a category, leaving the category out of the file is not enough.
Many stores keep data in extensions, for example extra fields for hazardous goods or unit prices. Check beforehand whether a plugin brings its own import profiles or fields. Which extensions have proven themselves is covered in the article Shopware 6 plugins.
In a relaunch, product URLs change. Without redirects you lose rankings and links. A CSV import creates new URLs, but no redirects from the old ones. Those need their own list.
No. The Import/Export module is included in Shopware 6. Extensions from the Shopware Store offer additional formats such as XML or Excel, schedules and more convenient mapping. For most cases the standard is enough.
UTF-8 encoding, semicolon as separator, quotation marks around text and a dot as decimal mark. The safest start is a file you exported from Shopware beforehand.
The documentation contains a field reference for orders but no guide. For moves from Shopware 5 or 6 the Migration Assistant is the better route, for other systems usually an import via API.
Yes, if you bring over the password hash and the method into the legacyPassword and legacyEncoder fields and Shopware supports that method. Otherwise customers reset their password once.
You do not. The import adds or overwrites. Deletions are done in the Administration or via API.
Many retailers do their imports themselves, and with this guide that is very doable for products, prices and customer lists. When variants, translations, passwords and redirects come together, the risk grows that mistakes only show up after go-live.
We help in three ways:
Write to us via the contact page and tell us which data you want to move and which system it comes from. We will tell you honestly whether you can manage it well yourself.

Besnik Bilalli is the founder and CEO of solution25, the Shopware agency from Bremen. Before founding the agency in 2017, he built his own online store and grew it to more than 500 orders a day. Since then he has been helping merchants with Shopware projects, from the first budget to ongoing operations. He’s happy to answer questions about this article personally: get in touch.
Responsible for content and updates: Besnik Bilalli. Last updated: November 2026. Next content review: February 2027.