An Amazon catalog flat file is a spreadsheet template that lets you create or change many product listings in one upload instead of editing them one at a time in Seller Central. For a brand with hundreds or thousands of SKUs, the file itself is the easy part. The hard part is the discipline around it: the right template, the right operation on every row, clean data from a single source, and a check of the processing report afterward. A brand that treats the flat file as a catalog operations process, not a one-off chore, rarely overwrites live content by accident. That is how we run it for the brands we manage on Amazon.
This guide walks through the full routine for US sellers: when to use a flat file, where to get the current template, how to fill it, how to upload it, and how to read the errors that come back.
A flat file is an Excel workbook (.xlsx) built by Amazon for a specific product type. Each row is a SKU. Each column is an attribute: title, brand, bullet points, images, dimensions, identifiers, variation settings and so on. Amazon’s own guide describes these inventory file templates as spreadsheets where you enter product data and upload listings through the Add Products via Upload page. Which attributes are required, and which are available, varies by product type.
Flat files earn their place in four situations:
For a single listing or a handful of edits, the Seller Central editor is faster. For daily price and quantity changes, a lighter inventory loader file is usually the better tool, because it only touches offer data. Flat files with hundreds of columns are for content and catalog structure.
If your brand sells as a vendor (1P) rather than through a seller account, catalog changes often run through a different workflow. Our Vendor Central management team handles that side; the rest of this guide covers seller-account (3P) uploads.
The wrong template is the most avoidable cause of failed uploads. The path in a US seller account looks like this:
Menu labels shift as Amazon updates Seller Central, so follow the on-screen wording if it differs slightly from the steps above.
Two habits matter more than the click path.
Download a new template for every upload session. Amazon changes fields, valid values and validation rules over time, and a template saved last quarter can fail today. Keep one master data source, not a folder of old workbooks.
Use the template that matches the marketplace. A US file uses US units and formats. Amazon.com and Amazon.de templates are not interchangeable, and a file built for another marketplace can fail on units, dates or accepted values.
To edit listings that already exist, start from a report of live data instead of a blank template. Seller Central’s inventory reports, such as the Category Listings Report and the Active Listings Report, give you current values to edit and send back. Starting from live data means you change only what you meant to change.
A generated template has several tabs. Three matter most.
Treat Valid Values as law. If a field has a controlled list, a near match such as “Navy Blue” where the list says “Blue” is a rejected value or a lost attribute.
This is where catalog teams earn their keep. Work through these in order.
The update_delete column tells Amazon what to do with each row. In practice there are three behaviors:
For existing listings, PartialUpdate is the safer default, with Update reserved for new listings or a deliberate full refresh in which every field is populated. Never leave the column blank on a row and hope for the best. Amazon’s current behavior for each value is documented in the Data Definitions tab and Seller Central help, so confirm it there before a large run.
Every product needs a valid product identifier (UPC, EAN, GTIN or ISBN, depending on the category) or an approved exemption. Match the identifier to the right product and variant. Mismatches between your identifier and the catalog’s existing record are behind some of the most common rejections, covered below.
Variation families fail when parent and child rows disagree. The rules in outline:
The template’s Check Your File feature, per Amazon’s guide, can identify common errors and configure variations. Variations also interact with Amazon’s policy, so it is worth reading our variation policy audit steps before you restructure a family that has sales history.
Titles, bullets and descriptions are not just conversion copy. Industry reporting on a February 26, 2026 change to Amazon’s browse tree guide says Amazon now leans on title, bullets and description to classify products, and that some listings fell out of alignment with their categories. If a bulk upload moves products into the wrong node, the copy may be the cause as much as the browse node field. Our step-by-step content optimization guide covers how to write copy that does both jobs.
Images deserve their own check. Flat files take image URLs, and Amazon fetches them at processing time, so a broken link or an undersized image can leave a listing without media. Our Amazon image requirements checklist lists what to confirm before the URLs go into the file.
Amazon publishes a guide to resolving common feed errors. These are the ones that most often stall brand uploads.
The barcode you submitted corresponds to an existing product in the catalog, and your attributes do not match that record. A frequent cause is a barcode that was reused. Amazon’s document gives apparel as an example, where manufacturers recycle barcodes after a few years. Check that the identifier really belongs to your product, then align your attributes with the correct record or correct the identifier.
The identifier is associated with two or more ASINs, and your title, brand, color or size match none of them. Clean up the attributes so they match one specific product, or use the correct identifier for that variant.
You are trying to list in a category that needs approval. The file is not the problem. Approval for the category has to be in place before the upload will work.
Required attributes are missing or invalid, so Amazon cannot create the product. Go back to the Data Definitions tab and complete every required field for that product type.
Remove HTML tags from the description field. Plain text only.
A working rule for any error: find the code and row in the report, fix the root cause in your master data, regenerate the file, and re-test on a small batch. Patching only the spreadsheet means the same error returns on the next upload.

Two changes affect how brands should plan.
First, Amazon has been moving programmatic listing management away from legacy formats. Its developer documentation announces the deprecation of XML and flat file listings feeds in the Feeds API in favor of JSON-based Listings Items APIs. That affects integrations that send flat files by API. Manual uploads through Seller Central are a separate path, and they are the workflow described here. If your team or a connected tool still submits flat files through the API, confirm with your integration owner that it has migrated.
Second, Amazon’s category structure and attribute requirements change from time to time, as the browse tree update above shows. That is the practical reason to download a fresh template before each upload and to schedule periodic catalog audits instead of assuming a clean upload stays clean.
Below a few dozen SKUs, a spreadsheet and a careful person can manage. Past a few hundred, with several marketplaces, regions and product lines, spreadsheets become the weak point. A workable structure for enterprise catalogs:
The costly pattern is uploading first and debugging afterward. A wrong price or an empty field on a top ASIN costs sales within hours, and no processing report undoes that.
Brands usually come to us when flat files have turned into a bottleneck: uploads stuck on the same errors, variation families that split, content that regressed after a bulk edit, or a catalog too large for one person to hold in their head. Our Seller Central management team runs catalog, inventory and compliance operations for 3P accounts, and our marketplace operations management covers the day-to-day work behind them. If catalog quality is part of a wider growth plan, Amazon Growth & Optimization pairs that operational work with content and advertising.

Because our creative studio produces photography, video and PDP content in-house, the images and copy that go into a flat file can come from the same team that manages the upload. See Creative & Content Production for what that covers. Prefer to start with a broader review? Our Strategic Consulting & Audits service looks at the catalog before anything is uploaded. You can reach us through thinknectar.com.
Most flat file problems are not file problems. They are data ownership problems. When no one owns the master record for a product, the spreadsheet becomes the record, and every Amazon template change turns into an emergency.
We push brands to treat the flat file as the last step of a process: clean source data, a mapping to Amazon’s fields, validation, a small test, then the full upload. It is slower on day one and much faster by month three.
The second thing we hold firmly is that catalog work and growth work belong together. A correctly classified, correctly varied listing with strong images and copy converts better and gives advertising a stronger page to send traffic to. That is why we manage operations, creative and media under one roof, and why our analytics platform, iDerive, puts catalog and performance data in the same view.
Dan Katona
No. You can add products one by one in Seller Central. Flat files make sense once you are creating or editing many listings, or when you need to repair variation families or run a catalog-wide content change.
No, download a new template each time. Amazon changes fields, valid values and validation rules, so a saved workbook can reject rows that worked before. Keep your data in a master source and move it into the fresh template.
Open Monitor Upload Status in Seller Central and download the processing report. It lists errors by row and code. Match each code to its cause, such as 8541 for a product ID that conflicts with an existing ASIN, then fix the underlying data and re-test with a small batch.
Amazon recommends testing with 15 or fewer products and uploading the remainder after a successful test. Include a variation parent and a few edge cases so the test shows the problems a full run would hit.
Amazon has deprecated XML and flat file listings feeds in the Feeds API and points developers to JSON-based Listings Items APIs. Manual uploads in Seller Central are a different workflow. If an integration submits flat files by API, ask whoever maintains it about migration.