Cloudflare yesterday announced its new Markdown for Agents feature, which serves machine-friendly versions of web content alongside traditional human-facing pages.

• Cloudflare described the update as a response to the rise of AI crawlers and agentic browsing.
• When a client requests text/markdown, Cloudflare fetches the HTML from the origin server, converts it at the edge, and returns a Markdown version.
• The response also includes a token estimate header intended to help developers manage context windows.
• Early reactions focused on the efficiency gains, as well as the broader implications of serving alternate representations of web content.

What’s happening. Cloudflare, which powers roughly 20% of the web, said Markdown for Agents uses standard HTTP content negotiation. If a client sends an Accept: text/markdown header, Cloudflare converts the HTML response on the fly and returns Markdown. The response includes Vary: accept, so caches store separate variants.

• Cloudflare positioned the opt-in feature as part of a shift in how content is discovered and consumed, with AI crawlers and agents benefiting from structured, lower-overhead text.
• Markdown can cut token usage by up to 80% compared to HTML, Cloudflare said.

Security concern. SEO consultant David McSweeney said Cloudflare’s Markdown for Agents feature could make AI cloaking trivial because the Accept: text/markdown header is forwarded to origin servers, effectively signaling that the request is from an AI agent.

• A standard request returns normal content, while a Markdown request can trigger a different HTML response that Cloudflare then converts and delivers to the AI, McSweeney showed on LinkedIn.
• The concern: sites could inject hidden instructions, altered product data, or other machine-only content, creating a “shadow web” for bots unless the header is stripped before reaching the origin.

Google and Bing’s markdown smackdown. Recent comments from Google and Microsoft representatives discourage publishers from creating separate markdown pages for large language models. Google’s John Mueller said:

• “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?”

And Microsoft’s Fabrice Canel said:

• “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 !”

• Cloudflare’s feature doesn’t create a second URL. However, it generates different representations based on request headers.

The case against markdown. Technical SEO consultant Jono Alderson said that once a machine-specific representation exists, platforms must decide whether to trust it, verify it against the human-facing version, or ignore it:

• “When you flatten a page into markdown, you don’t just remove clutter. You remove judgment, and you remove context.”
• “The moment you publish a machine-only representation of a page, you’ve created a second candidate version of reality. It doesn’t matter if you promise it’s generated from the same source or swear that it’s ‘the same content’. From the outside, a system now sees two representations and has to decide which one actually reflects the page.”

Dig deeper. Why LLM-only pages aren’t the answer to AI search

Why we care. Cloudflare’s move could make AI ingestion cheaper and cleaner. But could it be considered cloaking if you’re serving different content to humans and crawlers? To be continued…

Microsoft today launched AI Performance in Bing Webmaster Tools in beta. AI Performance lets you see where, and how often, your content is cited in AI-generated answers across Microsoft Copilot, Bing’s AI summaries, and select partner integrations, the company said.

• AI Performance in Bing Webmaster Tools shows which URLs are cited, which queries trigger those citations, and how citation activity changes over time.
• Search Engine Land first reported on Jan. 27 that Microsoft was testing the AI Performance report.

What’s new. AI Performance is a new, dedicated dashboard inside Bing Webmaster Tools. It tracks citation visibility across supported AI surfaces. Instead of measuring clicks or rankings, it shows whether your content is used to ground AI-generated answers.

• Microsoft framed the launch as an early step toward Generative Engine Optimization (GEO) tooling, designed to help publishers understand how their content shows up in AI-driven discovery.

What it looks like. Microsoft shared this image of AI Performance in Bing Webmaster Tools:

What the dashboard shows. The AI Performance dashboard introduces metrics focused specifically on AI citations:

• Total citations: How many times a site is cited as a source in AI-generated answers during a selected period.
• Average cited pages: The daily average number of unique URLs from a site referenced across AI experiences.
• Grounding queries: Sample query phrases AI systems used to retrieve and cite publisher content.
• Page-level citation activity: Citation counts by URL, highlighting which pages are referenced most often.
• Visibility trends over time: A timeline view showing how citation activity rises or falls across AI experiences.

These metrics only reflect citation frequency. They don’t indicate ranking, prominence, or how a page contributed to a specific AI answer.

Why we care. It’s good to know where and how your content gets cited, but Bing Webmaster Tools still won’t reveal how those citations translate into clicks, traffic, or any real business outcome. Without click data, publishers still can’t tell if AI visibility delivers value.

How to use it. Microsoft said publishers can use the data to:

• Confirm which pages are already cited in AI answers.
• Identify topics that consistently appear across AI-generated responses.
• Improve clarity, structure, and completeness on indexed pages that are cited less often.

The guidance mirrors familiar best practices: clear headings, evidence-backed claims, current information, and consistent entity representation across formats.

What’s next. Microsoft said it plans to “improve inclusion, attribution, and visibility across both search results and AI experiences,” and continue to “evolve these capabilities.”

Microsoft’s announcement. Introducing AI Performance in Bing Webmaster Tools Public Preview 

Eighty million people use Reddit search every week, Reddit said on its Q4 2025 earnings call last week. The increase followed a major change: Reddit merged its core search with its AI-powered Reddit Answers and began positioning the platform as a place where users can start — and finish — their searches.

• Executives framed the move as a response to changing behavior. People are increasingly researching products and making decisions by asking questions within communities rather than relying solely on traditional search engines.
• Reddit is betting it can keep more of that intent on-platform, rather than acting mainly as a source of links for elsewhere.

Why we care. Reddit is becoming a place where people start — and complete — their searches without ever touching Google. For brands, that means visibility on Reddit now matters as much as ranking in traditional and AI search for many buying decisions.

Reddit’s search ambitions. CEO Steve Huffman said Reddit made “significant progress” in Q4 by unifying keyword search with Reddit Answers, its AI-driven Q&A experience. Users can now move between standard search results and AI answers in a single interface, with Answers also appearing directly inside search results.

• “Reddit is already where people go to find things,” Huffman said, adding the company wants to become an “end-to-end search destination.”
• More than 80 million people searched Reddit weekly in Q4, up from 60 million a year earlier, as users increasingly come to the platform to research topics — not just scroll feeds or click through from Google.

Reddit Answers is growing. Reddit Answers is driving much of that growth. Huffman said Answers queries jumped from about 1 million a year ago to 15 million in Q4, while overall search usage rose sharply in parallel.

• He said Answers performs best for open-ended questions—what to buy, watch, or try—where people want multiple perspectives instead of a single factual answer. Those queries align naturally with Reddit’s community-driven discussions.
• Reddit is also expanding Answers beyond text. Huffman said the company is piloting “dynamic agentic search results” that include media formats, signaling a more interactive and immersive search experience ahead.

Search is a ‘big one’ for Reddit. Huffman said the company is testing new app layouts that give search prominent placement, including versions with a large, always-visible search bar at the top of the home screen.

• COO Jennifer Wong said search and Answers represent a major opportunity, even though monetization remains early on some surfaces.
• Wong described Reddit search behavior as “incremental and additive” to existing engagement and often tied to high-intent moments, such as researching purchases or comparing options.

AI answers make Reddit more important. Huffman also linked Reddit’s search push to its partnerships with Google and OpenAI. He said Reddit content is now the most-cited source in AI-generated answers, highlighting the platform’s growing influence on how people find information.

• Reddit sees AI summaries as an opportunity — to move users from AI answers into Reddit communities, where they can read discussions, ask follow-up questions, and participate.
• If someone asks “what the best speaker is,” he said, Reddit wants users to discover not just a summary, but the community where real people are actively debating the topic.

Reddit earnings. Reddit Reports Fourth Quarter and Full Year 2025 Results; Announces $1 Billion Share Repurchase Program

Google’s AI Mode isn’t more likely to cite content that appears “above the fold,” according to a study from SALT.agency, a technical SEO and content agency.

• After analyzing more than 2,000 URLs cited in AI Mode responses, researchers found no correlation between how high text appears on a page and whether Google’s AI selects it for citation.

Pixel depth doesn’t matter. AI Mode cited text from across entire pages, including content buried thousands of pixels down.

• Citation depth showed no meaningful relationship to visibility.
• Average depth varied by vertical, from about 2,400 pixels in travel to 4,600 pixels in SaaS, with many citations far below the traditional “above the fold” area.

Page layout affects depth, not visibility. Templates and design choices influenced how far down the cited text appeared, but not whether it was cited.

• Pages with large hero images or narrative layouts pushed cited text deeper, while simpler blog or FAQ-style pages surfaced citations earlier.
• No layout type showed a visibility advantage in AI Mode.

Descriptive subheadings matter. One consistent pattern emerged: AI Mode frequently highlighted a subheading and the sentence that followed it.

• This suggests Google uses heading structures to navigate content, then samples opening lines to assess relevance, behavior consistent with long-standing search practices, according to SALT.

What Google is likely doing. SALT believes AI Mode relies on the same fragment indexing technology Google has used for years. Pages are broken into sections, and the most relevant fragment is retrieved regardless of where it appears on the page.

What they’re saying. While the study examined only one structural factor and one AI model, the takeaway is clear: there’s no magic formula for AI Mode visibility. Dan Taylor, partner and head of innovation (organic and AI) at SALT.agency, said:

• “Our study confirms that there is no magic template or formula for increased visibility in AI Mode responses – and that AI Mode is not more likely to cite text from ‘above the fold.’ Instead, the best approach mirrors what’s worked in search for years: create well-structured, authoritative content that genuinely addresses the needs of your ideal customers.
• “…the data clearly debunks the idea that where the information sits within a page has an impact on whether it will be cited.”

Why we care. The findings challenge the idea that AI-specific templates or rigid page structures drive better AI Mode visibility. Chasing “AI-optimized” layouts may distract from work that actually matters.

About the research. SALT analyzed 2,318 unique URLs cited in AI Mode responses for high-value queries across travel, ecommerce, and SaaS. Using a Chrome bookmarklet and a 1920×1080 viewport, researchers recorded the vertical pixel position of the first highlighted character in each AI-cited fragment. They also cataloged layouts and elements, such as hero sections, FAQs, accordions, and tables of contents.

The study. Research: Does Structuring Your Content Improve the Chances of AI Mode Surfacing?