Google & Bing Advise Against Separate LLM Markdown Pages

```json
{
  "alt": "Close-up of blue computer code on a screen with blurred background.",
  "caption": "A glimpse into the digital world: code streams across a screen, shaping the virtual landscape.",
  "description": "The image shows a close-up view of blue computer code on a display, focusing on various 'if' conditions. The background is blurred, emphasizing the layered digital scripts. This scene captures the essence of software development and programming logic, highlighting keywords like 'RT_EVENT' and 'RT_CHAIN_EVENT'. It's a visual representation of the intricate and critical work done by developers in coding environments."
}
```

I’ve been following the lively debate around creating separate markdown pages for LLMs, and it appears that both Google and Bing are advising against this approach.

Recently, I noticed that representatives from Google Search and Bing Search have specifically recommended not to create separate markdown (.md) pages designed exclusively for LLMs. This practice involves presenting different content to the LLMs compared to what users see, which can be considered a form of cloaking—a direct violation of Google’s policies.

The question arose when Lily Ray inquired on Bluesky about the prevalence of creating markdown or JSON pages targeted at bots.

  • “Not sure if you can answer, but starting to hear a lot about creating separate markdown / JSON pages for LLMs and serving those URLs to bots.”

Google’s stance, as explained by John Mueller, is clear. He replied to Lily’s query saying that LLMs have always interacted with standard web pages and don’t require separate markdown pages.

  • “I’m not aware of anything in that regard. In my POV, LLMs have trained on—read & parsed—normal web pages since the beginning, it seems a given that they have no problems dealing with HTML. Why would they want to see a page that no user sees? And, if they check for equivalence, why not use HTML?”

John Mueller even criticized the whole idea, stating:

  • “Converting pages to markdown is such a stupid idea. Did you know LLMs can read images? WHY NOT TURN YOUR WHOLE SITE INTO AN IMAGE?” Of course, converting your entire site to a markdown format is an extreme measure.

I’ve collected many of John Mueller’s remarks on this topic, which you can find here.

Bing’s perspective is shared by Fabrice Canel from Microsoft Bing, who suggested that creating duplicate, non-user content isn’t effective.

  • “Lily: really want to double crawl load? We’ll crawl anyway to check similarity. Non-user versions (crawlable AJAX and like) are often neglected, broken. Humans eyes help fixing people and bot-viewed content. We like Schema in pages. AI makes us great at understanding web pages. Less is more in SEO!”

Why this matters to us: Many of us are tempted by shortcuts to improve search engine performance. Yet, these shortcuts often backfire or yield short-lived benefits. As Lily Ray remarked on LinkedIn, managing duplicate and differing content for bots violates established search engine policies.

Lily Ray’s thoughts on this are clear:

  • “I’ve had concerns the entire time about managing duplicate content and serving different content to crawlers than to humans, which I understand might be useful for AI search but directly violates search engines’ longstanding policies about this (basically cloaking).”

Inspired by this post on Search Engine Land.


crushpress.ai community screenshot

FAQs

Do Google and Bing recommend creating separate markdown pages for LLMs?

No. The post says representatives from both Google Search and Bing Search have advised against creating separate markdown pages designed only for LLMs.

Why can separate LLM markdown pages be risky for SEO?

The article explains that serving content to LLMs that differs from what users see can be considered cloaking. It also notes that this may violate established search engine policies.

What did John Mueller say about LLMs reading normal web pages?

John Mueller said LLMs have long read and parsed normal web pages and do not need separate markdown pages. He questioned why an LLM should see a page that no user sees.

What was Bing's view on duplicate non-user content?

Fabrice Canel from Microsoft Bing suggested that duplicate non-user versions add crawl load and are often neglected or broken. He also said Bing likes Schema in pages and that less is more in SEO.

What approach does the article imply is better than separate markdown pages for bots?

The post points toward keeping content in standard user-visible HTML and using Schema in pages. It warns that shortcuts involving duplicate or differing bot content can backfire.

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *