Spreadsheet workflow
How to find and fix common NIS file errors
Validate structure first, then review content and image links without changing unrelated columns.
Published 2026-10-03 · 8 minute read · By NoonFixer
Think in two layers
A large import becomes easier to diagnose when you distinguish row-level spreadsheet errors from decisions that belong in Seller Lab. The goal is not to change everything; it is to identify the smallest safe correction for each product.
Check structure before content
Open the file and confirm that it is the intended export or current template. Preserve sheet names, header rows, and required columns. A clean-looking spreadsheet can fail if it was saved from the wrong sheet or if headers were renamed.
Common structural flags:
- Missing SKU.
- Duplicate SKU rows in the same file.
- Missing English or Arabic title fields.
- No family, product type, subtype, or category value.
- Image links that are not HTTP or HTTPS.
Separate safe fixes from review decisions
A missing field is not always safe to auto-fill. A title may be drafted from verified product facts, but category and brand choices should match the current portal suggestion and your authorization to use that brand. Unknown values belong in a manual queue — never in a guessed cell.
Treat image errors as source problems
Replacing a filename extension does not convert an image. If a URL points to WebP, an inaccessible page, or a small source, prepare a real JPG or PNG file, host it at a stable public URL, and replace the link.
Work in controlled batches
Start with 20 products while validating your workflow. When the output is predictable, move to 50 or 100. The Bulk Issue Fixer keeps automatic title fills separate from category, brand, and image work, so each batch stays auditable.
Before the final upload
- Run the NIS validator.
- Resolve every error and review each warning.
- Compare the output columns with the latest official template.
- Keep a copy of the original export.
- Upload a small verified batch before a full catalog change.
Frequently asked questions
The 15-minute triage routine
When an import fails, do this before anything else: (1) open the import report and sort by error type, not by row; (2) fix the most frequent error class first — one systematic fix often clears hundreds of rows; (3) re-validate the file; (4) re-upload only failed rows. Random row-by-row fixing is how a 15-minute job becomes a 3-day job.
Building a personal error log
Keep a simple log: date, file, error text, root cause, fix. After a month you will own a checklist of your recurring mistakes — wrong date formats from one supplier, WebP links from one source, a column your VA keeps renaming. Prevention beats diagnosis every time.
When to ask for help
If the same rows fail twice with the same error after a correct fix, stop re-uploading and escalate: check Seller Lab notices for template changes, verify the category path still exists in the taxonomy, and confirm brand authorization. Repeated identical failures are usually a rule change, not a typo.
Automating the boring parts
Some checks should never be manual: duplicate-SKU detection, header-name verification, image-URL reachability, required-field presence. Build these into a pre-upload script or use the NIS Validator every single time — human eyes are for judgment calls (is this category right?), not for counting columns. Automate the mechanical; reserve attention for the meaningful.
Training your team on NIS
If someone else touches your files, they need three things: the current official template (not last year’s), the validator run as a mandatory step, and permission to stop and ask when a category is unclear. Most catalog damage is done by confident guessing — make “I don’t know, let’s check” the culturally rewarded answer.
The errors that look fixed but aren’t
Beware the false fix: an image URL that loads for you but not for the importer (signed-in session), a category that validates structurally but does not exist in the current taxonomy, a duplicate SKU removed from the file but still live in the catalog. After every “fix”, verify from the importer’s perspective — private window for URLs, current taxonomy for categories, Seller Lab for catalog state — not just from your spreadsheet.
Should I delete columns that look unused?
No. Preserve the current official template structure unless Noon’s instructions explicitly tell you to remove a field.
Can a valid spreadsheet still be rejected?
Yes. Structural validation cannot confirm every category rule, policy requirement, brand decision, or manual quality review.
How many products should I include per import?
Start small — 20 products — until your workflow is predictable, then scale to 50 or 100 per batch.