> For the complete documentation index, see [llms.txt](https://docs.sageseo.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.sageseo.ai/reference/article-enrichment.md).

# Article Enrichment

After Sage writes an article, a second AI pass goes back through the finished content and adds **structured supplemental assets** — Key Takeaways, FAQs, internal links, optional charts. These aren't decoration. They're the elements that actually drive citations from Google + AI search engines.

## Why structured content wins

Google's AI Overviews, ChatGPT, Perplexity, and Bing Copilot all do the same thing when answering a user's question: they scan candidate pages, pull out the most extractable answer, and cite it.

A wall of plain text loses every time. A page with:

* A clearly-labeled "Key Takeaways" callout near the top
* One-sentence direct answers under question headings
* A FAQ section with explicit Q\&A pairs
* Structured FAQPage schema markup

...is what gets cited. Not because it's "better content" in some abstract sense, but because it's literally easier for AI systems to extract from.

Sage's enrichment exists to give every article these extractability surfaces by default — so you don't have to think about them on a per-article basis.

## What enrichment adds

### Key Takeaways

A blue callout box at the top of the article with 3–5 short bullets, each capturing one main idea in plain language. This is the single most-cited element on every page — both by Google's AI Overviews and by ChatGPT/Perplexity.

### FAQ section

4–5 question-and-answer pairs at the bottom of the article, formatted with explicit H3 questions and 1–3 sentence answers. Sage attaches structured FAQPage schema so Google can render the FAQs directly in search results.

### Internal links

2–3 suggested links to other pages on **your own site** (pulled from your sitemap), placed at relevant points in the article body. Internal linking is one of Google's strongest ranking signals — pages that connect well to your domain's other content rank higher than orphaned articles.

### Charts + diagrams

When the article cites comparable numbers, statistics, or competing options, Sage adds an embedded chart visualization at the natural insertion point. Charts increase dwell time + social shareability and signal "real content" vs "AI fluff" to readers.

### Pull quotes

The most memorable sentences from the article body. Sage surfaces these in the preview so you can decide whether to manually pull-quote them in your editor — they're not auto-injected to avoid duplicating content in the published page.

## What enrichment doesn't do

* **Doesn't rewrite the body** — the article writer's output is the canonical body; enrichment only adds to it
* **Doesn't invent stats or facts** — every chart datapoint, every Q\&A answer comes from content the writer already produced
* **Doesn't link to external sites** — only internal links from your sitemap
* **Doesn't replace good content** — enrichment polishes; it can't save a thin article

## When enrichment runs

* **Automatic** — every article generated through Sage (autopilot OR manual generation) gets enriched as part of the standard pipeline
* **Re-enrichment** — if you edit an article body significantly, you can trigger a fresh enrichment pass to re-derive takeaways/FAQs from the new content (coming soon)

## Why this matters more than people think

Most AI writing tools generate the body and stop. The output ranks badly because it's hard to extract from. Sage's enrichment is what turns a generated article into a publishable + citable piece — the difference between "ranks on page 5 with zero clicks" and "shows up as a featured snippet on page 1."

It's also what makes AI search citations possible. When ChatGPT or Google's AI Overview answers a query and shows source links, it cites pages it could extract from cleanly. Enriched pages are the candidates.

## Related

* [How Sage Publishes Articles](/reference/publishing-mechanics.md) — enrichment is woven into the published HTML
* [URL Inspection](/reference/url-inspection.md) — once published, this tracks whether Google accepts the article into its index