Latest Technical SEO News - Page 11
Crawling changes, algorithm updates, rendering fixes, and structured data developments as they happen.
-
Vercel lets in-page AI agents call your site's tools as the user
Vercel's mcp-handler 2.2.0 serves a WebMCP script that registers allowlisted MCP tools in the page and runs each call with the visitor's session cookies.
-
AI visibility swings of 20 points can be noise at 10 runs a prompt
Mike King of iPullRank publishes margins of error by run count, explains why vendors disagree on model mode and prompt phrasing, and lists methodology questions.
-
Cloudflare's AI training block now stops Googlebot too
Since September 15, Cloudflare's Training block also shuts out Googlebot, Bingbot and Applebot. Disallow AI Training refuses training and keeps search.
-
JavaScript error pages can make Google index another site instead
A Reddit user saw Google pick a casino page as canonical for their pages. Mueller says an indexed JavaScript error page may explain it and suggests hourly checks.
-
Your September visibility drop may be a rank tracker losing data
Rank tracker Nozzle lost about 80% of its Google data in mid-September and Sistrix's collection slowed, so check Search Console before trusting a visibility drop.
-
Google says AdSense crawler rules also affect Ad Manager
Google's crawler docs now say Mediapartners-Google rules affect several ad products, and the crawler ignores the * group, so only a named robots.txt group stops it.
-
Charging AI agents per page works, but no search crawler can pay
Suganthan Mohanadasan's x402 demo charges AI agents one cent a page. Only agents Mohanadasan runs have paid, in testnet money, and Googlebot cannot pay at all.
-
Google's AI agent dropped local firms with vague pages in a test
A Reddit tester says Google's agent kept home service firms stating fixed quotes or same-day service, and the fixes are in page copy and LocalBusiness markup.
-
ChatGPT's default model cites brand sites far less than Thinking
Writesonic ran 50 prompts through two ChatGPT models. Brand websites got 8% of the default model's citations and 56% of Thinking's, and sources barely overlapped.
-
AI answers cite retailers' size guides, support and policy pages
Aleyda Solis reviewed Semrush and Similarweb data for 25 US retailers and found AI answers cite size guides, help articles and policies alongside product pages.
-
Google crawled a news site less as its soft 404 count climbed
A Search Engine Land case study traces a 90% traffic loss at a crypto news site to a botched migration, then finds soft 404s growing on its other country sites.
-
Google deindexed a 1,800-post blog that passed technical checks
A gaming blog's 1,015 pages sat at Crawled - currently not indexed with blank canonicals, and a Screaming Frog crawl found exploit scripts and download gate code.
-
Robots.txt only stops AI crawlers that choose to obey it
Reverse DNS for Googlebot, OpenAI's published IP lists, and CDN rate limits deal with the AI bots a Reddit thread found hitting parameter URLs despite robots.txt.
-
AI Overviews and ChatGPT usually criticize different brands
BrightEdge compared negative brand mentions in Google AI Overviews and ChatGPT: Google leans on lawsuits and recalls, ChatGPT on product limits and compatibility.
-
Adding schema did not raise AI citations in an Ahrefs study
Ahrefs tracked 1,885 pages that added JSON-LD against matched controls and found no citation lift in ChatGPT or AI Mode, plus a small dip in AI Overviews.
-
Google may reorder results using hovers and scrolls on the SERP
SALT.agency describes Glue, a Google log of hovers, scrolls and viewing time on SERP features that feeds a popularity score used to reorder ranked results.
-
Fix soft 404s on empty category pages by how long stock is gone
Search Console flags empty category pages as soft 404s. Added content, a 302 or noindex each fits a different stockout length, and a canonical fits none.
-
Lighthouse now grades pages for AI agents on accessibility and CLS
Lighthouse 13.3's Agentic Browsing category checks the accessibility tree, CLS, WebMCP form annotations and llms.txt, and sites without AI features still pass.
-
Migrations that merge pages need a staging crawl before launch
A Reddit thread on moving a custom CMS store to Shopify shows why merged pages and Shopify's own canonical tags call for a full staging crawl before launch.
-
AI agents built on Cloudflare may never run your JavaScript
Cloudflare's Project Think SDK lists a headless browser as an add-on tier that agents may never use, so schema and content injected by JavaScript can go unseen.