Amazon Error 90122: How to Remove HTML from Your Product Data

Applies to inventory flat file uploads via Seller Central · Updated July 2026

Error 90122 means Amazon detected HTML tags in one of your product data fields — usually the description, sometimes bullet points or even the title. Amazon's detail pages don't accept HTML in listing data, and rows that contain it get rejected.

The confusing part: you probably never typed a single tag. The HTML almost always rides in invisibly with copied text.

Where the hidden HTML comes from

Source	What it smuggles in
Copy-paste from your website / Shopify / eBay listing	`<div>`, `<span>`, `<strong>`, inline `style=` attributes
Word or Google Docs drafts	Proprietary markup that converts to tags on paste
Old listings migrated from another platform	`<br>` and `<p>` from the platform's rich-text editor
AI-generated copy	Literal `<br>` or `\n` line-break artifacts
Feed/PIM tools with rich-text fields	Formatting silently serialized as HTML on export

Step-by-step fix

1. Find every affected field

The processing report names the SKU, but not always every dirty field. In your flat file, run a search (Ctrl+F in Excel) for < — the opening angle bracket. Check the columns that hold free text:

product_description — the usual offender
bullet_point1 through bullet_point5
item_name
generic_keywords (rare, but paste accidents happen)

2. Strip the tags — the plain-text round trip

For a handful of SKUs, the most reliable manual method:

Copy the field's text into a plain-text editor — Notepad on Windows, TextEdit in plain-text mode on Mac. This flattens all formatting.
Delete any visible tags that remain (<br>, <b>, <li>…) — the editor only removed formatting, not literal tag text.
Paste the clean text back into the cell. Don't route it through Word or a web editor on the way.

For hundreds of SKUs, do it in-sheet: Excel's Find & Replace with wildcards (find <*>, replace with nothing) removes well-formed tags in one pass. Spot-check afterwards — nested or broken tags can leave fragments like a stray >.

3. Replace formatting with plain-text structure

Sellers usually had the HTML there for a reason — paragraph breaks, emphasis, bullet lists. Rebuild the same readability without tags:

Line breaks: in Excel, use Alt+Enter inside the cell instead of <br>.
Bullets in the description: plain dashes or the • character.
Emphasis: capitalize a short phrase sparingly — there is no bold in plain text.
Real bullet lists belong in bullet_point1–5, not simulated inside the description.

4. Re-upload with `PartialUpdate`

Set update_delete to PartialUpdate so you only overwrite the text fields you cleaned, then read the new processing report and confirm the 90122 rows are gone.

"But other listings use HTML and they're fine." You'll see legacy listings with rendered formatting, and some guides claim <br> still slips through in descriptions. Maybe — but enforcement has tightened year over year, and a feed that relies on tolerated tags breaks on Amazon's schedule, not yours. If you want rich formatting, the supported path is A+ Content (Seller Central → Advertising → A+ Content Manager, requires Brand Registry), which replaces the plain description on the detail page.

Preventing the next 90122

Draft listing copy in plain text from the start; treat Word/Docs as a no-go for final copy.
Add a pre-upload habit: Ctrl+F for < across the sheet. Two seconds, catches nearly everything.
If a feed tool builds your files, turn on its HTML-stripping option and inspect its raw export once — rich-text fields are the usual leak.

Or skip the spreadsheet archaeology

Upload your flat file to FlatFileFix — it locates every field with embedded HTML across all your SKUs, strips the tags while preserving the text, and gives you a clean corrected file. Diagnosis is free.

Fix my flat file →