I've developed a framework that combines authority metrics with semantic relevance to return much deeper insights into link value. seodepths.com/python-for-s...

Log file analysis shows llms.txt has limited impact on AI crawl behaviour. Learn why robots.txt still matters most and how to deploy llms.txt safely. seodepths.com/seo-research...

🔔 Did you know #OpenAI may keep a cached index of webpages? When a page is fresh or heavy on #JavaScript, #ChatGPT can generate answers from related cached pages. I built a small app using your OpenAI API to check the external_web_access endpoint Tests and context -> github.com/simodepth96/...

I'm not surprised at all. If you're based in the EU and have a proper look into your server logs, this is what you get. The idea per se is smart as LLMs may slice .md faster than .html but if it doesn't become a standard in the W3C then it lives in thin air

this goes right along with my theory that the AI obsession in *users* is often bc… 1. they were never educated (esp "ipad generation") & don't know how computers work 2. our tools and services are all SO AWFUL NOW, degraded at best, actively user-hostile at worst

Head to Skyline to attend an insightful session on ‘Technical SEO nightmares on client-side React web apps’ by @simodepth.bsky.social #BrightonSEO

A test on 7 URLs with missing meta descriptions showed: - Google and ChatGPT favoured H1S and structured text - Perplexity relied on cached metas. LLMs and AI search engines tend to mimic Google, preferring structured, self-contained content during retrieval. 👉🏻 seodepths.com/seo-research...

Gemini API Grounding might be the reference point to start optimising for fan-out search queries from #AIMode In the screenshot, the results of a few tweaks to the #Python from the cookbook repository ai.google.dev/gemini-api/d... 🧵