Cyrus S.'s Recent LinkedIn Posts

Google spent 20+ years fighting spam, refining algorithms, establishing trust and authority metrics, and generally protecting the integrity of its results Now it's *STILL* being undone by Top 10, self-serving listicles. After reading an excellent write-up by Dan Petrovic, been thinking about this A LOT Google clearly recognizes the problem. It's all over social media. The 2025 December Core Update reportedly cracked down on many SaaS providers who published self-serving listicles, as observed by Lily Ray But sites are still doing it. They're adding: • "Original research" (fake) • "Experience"(fake) • "Expert insights" (again, fake) And in an effort to get into LLM training data, these lists are being published on parasite sites, LinkedIn, Reddit, and anywhere you can imagine. And Google and AI companies are falling for it again and again. 🤦 (Aside from that, it's a poor user experience) What's going to happen? Google doesn't like to be embarrassed. Expect a few things. My predictions: 1. Google will add specific language to the Quality Rater Guidelines asking raters to explicitly downgrade self-serving listicles 2. At a certain point, Google will begin classifying more of these lists as spam 3. Authority and Trust metrics will get turned way up, as usually happens in these cases. (Unfortunately, this also means that high-authority sites will be more likely to get away with it.) None of this is going to happen in the next few weeks, but in the next few months? I wouldn't bet against it.

Looking forward to this. Great gathering of SEO folks + terrific location. Q: Since I've never been to Poland, what's the best food to eat? cc: WhitePress®

As "traditional" search becomes strange in the age of AI, SEOs should take greater ownership of search across the entire ecosystem - not only Google It's not like what we do doesn't influence all these other platforms. Search is everywhere.

Endorsed. Lee Foot has built a fantastic tool for internal linking. Find the most semantically relevant links between pages (first 1000 pages free)

My fortune cookie is oddly specific 🤔

The first "E" in E-E-A-T stands for "Experience." Can you fake it for Google rankings? Google's Quality Rater Guidelines (and its algorithms) are crafted to reward first-hand experience and expertise. The folks at CouponBirds did a breakdown of a competitor that's dominated the super competitive coupon space. The coupons look like they are curated by human editors, but when you peel back the curtain, it's all AI. Fake AI personas, fake AI everything. Is it fooling Google? Can it work as a long-term strategy, or will it crash and burn? You be the judge: https://lnkd.in/daMth69Y

Is the silo finally breaking? Does it need to break in order for the team to handle visibility right now? LLM visibility is shaped by far more than just your website. The equation now includes PR, expert mentions, reviews, product messaging, social, community, support — and everything else SEO might not fully control. So we asked experts: ⚡ Does team structure need to change for SEOs to succeed at LLM visibility? The answers weren’t really about job titles. They were about influence, alignment, and whether SEO is actually involved where brand perception gets shaped. Big thanks to Nitin Manchanda, James Brockbank, Jack Chambers-Ward, Duane Forrester, Cyrus S., Chris Long, Tess Voecks, Bengü Sarıca Dinçer, Pedro Dias, Katerina Fotiadi, and Rejoice Ojiaku 🔥 Would love to hear the honest answer: 👉 Does the structure need to change?

Google's "ABC" signals: The 3 fundamental signals of ranking One of my favorite documents from Google's antitrust trial, from an interview with Google Engineer HJ Kim • A for "Anchors" - A source page pointing to a target page (links) • B for "Body" - Terms in the document • C for "Clicks" - User interaction signals, e.g., "historically, how long a user stayed at a particular linked page before bouncing back to the SERP" ABC signals help Google determine Topicality - is the document relevant/engaging/useful? They discuss other signals discussed. Importantly, QUALITY (Q*), eDeepRank, and PageRank Favorite lines from this interview: 1. "There was a leak of Google documents which named certain components of Google's ranking system, but the documents don't go into specifics of the curves and thresholds." 2. "By User Side Data, Google's search engineers mean user interaction data, not the content/data that was created by users." 3. "Nowadays, people still complain about quality and AI makes it worse." Source: Trial Exhibit-PXR0356: U.S. and Plaintiff States v. Google LLC.

If you need another reason why you should prioritize optimizing for Google > ChatGPT (if you had to choose), our friends at SISTRIX lay out all the evidence for how Google/Gemini is: • Growing • Winning • Has huge structural cost + platform advantages OpenAI had the headstart, but now all the winds are at Google's back. Of course, there's no reason not to optimize for both :) Nice work Johannes Beus 🙌

Lots of tidbits in this study from Define Media Group from their publishing panel of *64-sites and over 1 billion clicks per quarter* - with some of the largest publishers on the planet 🌎 Key Findings: 1) AI Overviews have reduced clicks -42% across this portfolio 📉 2) But... clicks from Google Discover are UP 30% over the same timeframe 📈 3) Breaking News traffic up by 103% 4) News / Sports / Real Estate / Shopping trigger AI Overviews far less than other verticals Regardless, great data for publishers looking where to invest in content production + optimization strategies. Nice work Shahzad Abbas 👏 https://lnkd.in/g5Rd9sPT

Excited to Launch a different kind of newsletter for changing times in SEO/Marketing: "Zyppy Signal" Here's the scoop... • What It's About: Brand/AI/SEO - Marketing your brand through search engines and AI, because brand is the most important signal 🧭 • Who It's For: Everyone on the marketing team: CMOs, SEOs, Brand/Growth Marketers, and anyone with a website 🙋 • What To Expect: At least 1 new strategy every week 📋 Sign up here if you'd like to receive the first edition free: https://signal.zyppy.com/ Thank you for your attention to this important matter

Wow. Super impressed with the work Metehan Yeşilyurt is doing with Google Discover 👏 In a nutshell: he's reverse-engineering how Google Discovery works, using the underlying code Google exposes whenever Discovery runs on any device He's uncovered incredible details about: • time decay • which meta tags are processed • fact-checking • and dozens of other ranking considerations Bonus! Using these ranking factors, he's built AI-driven tools to evaluate how Discover could evaluate your content The "Discover pCTR Predictor" below analyzes potential Google Discover titles on Entity Density, Clickbait penalty, Freshness signals, and more Caveat: Nobody knows the exact weights Google uses, but these tools at least do an excellent job of showing you what Google Discover is considering. Finally, you should definitely subscribe to his newsletter. One of the smartest folks in the SEO / AI space I follow! More in the comments

The data is in. The most cited source in Google AI mode is... Google! Congrats? 👏 Second place goes to YouTube (also Google) 👏 👏 The study from Yulia Deda and Svitlana Tomko looked at 68,313 keywords and 1,321,398 citations While it's challenging that so many clicks aren't going to traditional publishers, the silver lining is that so many citation sources can be influenced by the marketing team: • YouTube (work those transcripts) • Facebook • Reddit • Amazon (reviews) • Instagram • Wikipedia (harder, but doable) For the full study, Google "Is Google stealing your clicks in AI Mode? (1.3M+ citations analyzed)" What's your view on citations in Google's AI mode?

Spicy Take 🌶️: Companies looking to hire you aren't actually shopping for SEO / GEO / AI SEO / Insert-name-here What they *really* want is to solve a business problem If you can increase sales + visibility + conversions across online surfaces, you can call it whatever you want In my experience, people who sell solutions > people who sell the process

Here's the truth, I replaced my entire SEO agency with 8 AI Agents Here's how: I never had an SEO agency

Google's new "AI-powered configuration" in Search Console is cool and all, but it *does not* give us what we all really want

Whenever I see domains tanking in the SERPs, I go back to what Liz Reid, Google's VP of Search, said in 2025. Wired ⚡ : Personal Perspectives, Expertise, Effort, Videos, UGC Tired 📉 : Content that says what everyone else knows Not a universal rule, but it holds up incredibly well

New research via AirOps shows data that marketers are likely still missing: 1) Confirmed: Ranking in Google for the keyword is highly correlated with getting cited in ChatGPT "55.8% of all cited pages ranked in the top 20 for at least one query" 2) Slightly less obvious: Fan-out queries create a *second* citation opportunity "32.9% of cited pages that appeared in any top-20 SERP were discovered only through fan-out" While this research is specific to ChatGPT, I wouldn't be surprised if we observed directionally similar trends with Google AI, where citations can drive meaningfully significant traffic. My 2¢: Most marketing/content teams are focused on #1 (ranking for the head keyword) and only starting to glance at #2 (fan-out queries). The challenge: Fan-out queries can grow exponentially. If you want to optimize for them, you need to collate/group them across lots and lots of prompts to find the most high-value content gaps in your AI strategy. Are you optimizing for fan-out queries? Why or why not?

New: "How Google Click Signals Drive Rankings and AI Answers" We've known for a long time that Google uses click data for evaluation, training data, and re-ranking. But did you know that click data also informs AI answers? And the influence of Google click data can easily surface on other platforms like ChatGPT? So if click data is so important to Google, we should understand: 1) What those click signals are 2) How Google uses them, and 3) Work to make our sites better for users in a way that Google wants to reward. Not only SEOs. Your entire marketing and engineering teams should understand this. Because even if Google didn't measure click signals, you've made your website better. I had 100% fun diving into clicks, goodClicks, lastLongestClicks, and more. Post in thread 🙏

New study of 15K URLs shows that articles receiving Major Rewrites (30-100% of content) vastly outperform minor/moderate updates Even minor content updates helped prevent decay, but content with no update performed the worst Interestingly, even "minor" rewrites, defined as 0-30% of content, still saw decay - although not as much Is this the cutoff for "lastSignificantUpdate" - found in the Google API Leak, which Shaun Anderson defines as "A timestamp (in seconds) indicating the last time the document underwent a substantial content change, distinguishing it from minor cosmetic edits."? Regardless, my 2¢ here is that content updates work to prevent ranking decay, but "significant" content updates may work much, much better Thoughts?

Do AI companies check if you have an LLMs txt file? Nope, certainly not by default! That's what Reboot found in a clever experiment ⤵️ They set up LLMs txt files, including links to uncrawled pages, then monitored the log files for visits After THREE MONTHS, neither the LLMs.txt files nor the new pages were ever visited by any AI bots - despite the AI crawlers visiting multiple pages on each domain Bravo for this fine SEO detective work 👏 https://lnkd.in/g66y7TsN My 2¢ - the LLMs txt file is, at this point, simply a page like any other. If it's not important to your human visitors, it's likely not important to AI/Search Engines.

Extremely well done "The State Of AI Search White Paper" by the Prosperity Media team For any CMO/Marketing Team/SEO wanting to learn how a 1st-rate Marketing Team is approaching AI Best part: 100% Free - no email required