> This page is for Product.

> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.feedonomics.com/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.feedonomics.com/_mcp/server.

# What enrichment does (and doesn't do)

> What enrichment generates — product copy, SEO, AEO, and channel fields — and what it deliberately does not do, including structured attributes and categorization.

Before you run your first enrichment, it helps to know exactly what the product does — and just as importantly, what it deliberately doesn't do. Clear expectations here save you review time later.

## What enrichment does

FDX Enrichment generates **language**: the human-readable content that sells your products and helps channels and answer engines understand them.

* **Product copy** — titles, short and long descriptions, feature bullets, care instructions.
* **SEO and discovery metadata** — meta descriptions, image alt tags, and discovery tags.
* **AEO content** — AI-ready descriptions and conversational Q\&A that answer engines can cite.
* **Channel-specific copy** — content shaped to the requirements of Google, Meta, TikTok, Amazon, eBay, and Walmart feeds.

For the complete list — every field a run can produce and what each one is for — see [Fields in an enrichment run](/product/enrichment/fields).

Everything it writes is grounded in your source data and governed by your brand guidelines. If your data says a jacket is water-resistant, enrichment can say so persuasively — but it will never promote that to "waterproof."

## What enrichment doesn't do

> **Enrichment writes copy, not structured attributes**
>
> FDX Enrichment generates language: titles, descriptions, feature bullets, Q\&A, and SEO text. It does not infer or fill structured attribute values like color, material, or size, and it does not assign products to taxonomy categories — that work is **attribution**, which is **not part of data enrichment today but is on the roadmap**. If a structured attribute is missing from your source data, enrichment won't invent it.

A few more boundaries worth knowing up front:

* **It doesn't overwrite anything.** All output lands in new, dedicated enrichment fields.
* **It doesn't guess at missing facts.** Gaps in your source data stay gaps — flagged, not filled with fiction.
* **It doesn't write meta keywords or search tags.** Search engines have ignored those for years, so we focus SEO output where it actually moves rankings: page titles and meta descriptions.

> **Output is US English only**
>
> Enriched content is generated in **US English only**. All output uses US English spelling, grammar, and conventions.

## Why the line matters

Copy and structured data are different jobs with different bars. Copy is judged on quality — is it accurate, consistent, on-brand? Structured attributes are judged on exactness — a wrong color value breaks filters and channel compliance. FDX Enrichment is built for the first job and holds itself to that quality bar on every run. You'll see exactly how in [Understanding your quality scores](/product/enrichment/scorecard).

Filling those structured values is **attribution**, and it's a distinct capability. **Attribution is not included in data enrichment today, and it is on the roadmap** — so plan on structured attribute values coming from your source systems. [Fields in an enrichment run](/product/enrichment/fields) draws the line field by field.

For where each field lands per destination, see [Where enriched fields go](/product/enrichment/field-mapping).