Did you know that Pinterest also uses an algorithm called NavBoost? We reverse engineered the app to uncover it. Discover the 263 signals that determine why your Pin gets selected over another think.resoneo.com/pinterest/

« Tailor Your Feed » : le fan-out Google Discover qui fait émerger les sites de niche Et pourquoi ça ouvre une troisième voie de visibilité aux petits sites, au-delà de l’affinité implicite et du Follow A lire sur Abondance: abondance.com/20260609-24088… @5eg Et un nouveau fan out!

For 3 months we decompiled Google's apps. 100K source files, read class by class. What we found isn't a better search engine. It's the blueprint for an OS for human intent👇 seg1795578.substack.com/p/wh…

🔥 For the first time, Google and Schema published how many domains use each Schema term. 5,545 terms. I ranked them by adoption. Here's what it shows ↓ 1/7

Ce que je retiens de cette annonce c’est que même Google considère que le GEO est un sous ensemble du SEO 😬

Google I/O GML : Gemini devient l'infrastructure. Le contrôle se déplace en amont ✨ #SEA : Fin des mots-clés, les DSA migrent vers AI Max en septembre. #SEO : Seuls 38% des liens AI Overviews viennent du top 10 organique. Notre décryptage complet 👇 resoneo.com/google-i-o-et-ma…

In modern AI search, language models act as re-rankers over results retrieved by traditional rank factors. Even in the absence of traditional ten blue links there's always a clear and measurable ordinal value to each brand mentioned in model's generative output. Here's how I test this: dejan.ai/optimizer/ Google's grounding pipeline, for instance, decomposes a query, retrieves ranked sources, then has the model select sentence-level snippets under a fixed budget. Ranking #1 buys you a larger share of grounding, though being selected is a separate problem [1]. A model's parametric memory carries its own relevance priors, and those priors are an emerging class of factor shaping which results get selected and surfaced. A brand the model already perceives as relevant for a topic is more likely to be grounded when supplied as a source [2], and these priors are measurable: you can rank brands by how deeply they're embedded in a model's associative structure [3]. To be clear about terminology: when I say model rank factors I mean model-side selection factors. They're distinct enough from Google's ranking signals that I've taxonomized them as alignment, substance, architecture, style, framing, and proof, and built a ranker that simulates the model's source-selection step to measure which of them actually move a page's standing [4]. My current focus is understanding this behaviour through systematic observation of inputs and outputs, probing models directly and tracking how associations shift over time [5]. Direct steering and white-box interpretability aren't available for closed-weight models like Gemini, GPT and Claude, so this black-box approach is the practical one. It's the same logic applied psychology, psychiatry and cognitive science already use. [1] SRO & Grounding Snippets dejan.ai/blog/sro-grounding-… [2] Primary Bias on Selection Rate dejan.ai/blog/sr/ [3] AI Brand Authority Index dejan.ai/blog/brands/ [4] Content Optimizer dejan.ai/optimizer/ [5] Beyond Rank Tracking dejan.ai/blog/beyond-rank-tr…

We pulled Chrome apart and found secret classification files. Will you find your site? Lists baked into the browser that decide what your site is and what AI is allowed to do with it 2M domains tested and a tool to read it live: think.resoneo.com/chrome-cla… cc @dejanseo

Gros spam russe sur les liens de redirection Youtube qui sont indexés : google.com/search?q=site:you…

Ever wondered what Google AI Mode can actually do behind the prompt box? We mapped the full list of tools and skills it runs on. Here is each one, and how it works Grammar = tool_name : function Before the colon sits the tool, after it the function it calls

No es la primera vez que me encuentro cosas de este estilo en AI Mode o las AI Overviews. Tiene pinta de ser Google tirando directamente del Shopping Graph para sacar entidades y normalizar el grounding del Web Graph. ¿Lo habéis visto alguna vez? #Leak

Frandroid plafonne à 5 pipelines #GoogleDiscover. Le Monde en touche 12. ➡️Ce n'est pas une question de SEO, c'est une question de format éditorial. La synthèse de notre série : 7 profils d'éditeur, leurs leviers, leurs limites. 1492vision.substack.com/p/re…

What Microsoft is implicitly saying is that the value of search engines is no longer to send traffic but to generate the best answer. They still need open web and publishers, but more as providers of data and knowledge than as click destinations. This model won’t last long term…

Tomorrow's AI phone, seen from inside a Google APK ==> what the intelligent phone will be in 12 to 18 months think.resoneo.com/phone-futu…

Petite question aux utilisateurs de Claude Code, vous avez déjà testé ca ? github.com/rtk-ai/rtk Ca fonctionne bien ? Ca compresse aussi lorsque Claude utilise Read, Grep? pas seulement Bash ? Merci pour vos feedbacks ^^

Google Preferred Sources: we scraped ~10K domains per country (US, UK, DE, ES, FR, AU). See what Google suggests as “trusted sources” : think.resoneo.com/google-pre… Useful for PR targeting, link building, etc :)

What's the difference between OAI-SearchBot and ChatGPT-User? The whole market gets it wrong, and OpenAI's docs don't help... ==> linkedin.com/feed/update/urn…

Ibou Explorer, 6 mois plus tard open.substack.com/pub/ibouio…