Frederik Drost

Big announcement! After 16 years in SEO, today I'm launching something I've been building toward for a long time. 🥳 Introducing Algorythmic: my new SEO and AI search consultancy. For the past decade+, I've led award-winning SEO teams at agencies. I've spoken at ~100 conferences. I've spent thousands of hours analyzing Google algorithm updates, studying E-E-A-T, and more recently, figuring out how brands can show up well in AI-driven search. Now I'm channeling all of that into something of my own. Through Algorythmic, I'll be selectively working with brands 1:1 on SEO consulting, AI search optimization (AEO/GEO), E-E-A-T strategy, content quality audits, Google Discover, algorithm update recovery, hourly training, and more. I'm also excited to announce that as part of this launch, I will be continuing my role as VP of SEO & AI Search at Amsive, where I still oversee an incredible team of 30+ SEO experts (who won "Best Enterprise SEO Team" in 2025, according to the Search Engine Land awards!). Algorythmic will allow me to take on solo projects that are a strong personal fit for my skills and experience. Check out the comments for the link to my new site. BTW, if you're curious about the name Algorythmic: it's a mashup of "algorithm" and "rhythm." If you know me, you know those are the two things that have defined my entire life (especially 'rhythm'). The full origin story (involving a fictional SEO-themed deli my Amsive team created in 2019) is on the blog - link also in comments. I'm being very intentional about the work I take on through Algorythmic. If you think we'd be a good fit, I'd love to hear from you. Check out the new site and get in touch!

@MalteLandwehr @Reddit 😵😵

Will @Reddit ever tackle their blatant spam problem? This user has written the same question already 100+ times.

@lilyraynyc @CyrusShepard - Thought you would appreciate this XD

@KevinNaughtonJr Takes a real man to admit mistakes.. that is for sharing we can all learn from this !

I was fired from Anthropic today. I was the engineer responsible for shipping the latest dev/claude-code npm package. Wanting to improve the debugging experience for the team, I decided to include source maps in the release. This resulted in our entire internal codebase being publicly exposed including thousands of files with every agent command, all system prompts, the complete query engine, Undercover Mode, Bypass Permissions Mode, and our internal telemetry configuration. I take full responsibility. I genuinely believed the safeguards Claude Code had built for me would be adequate and it was a serious miscalculation on my part. My actions have unintentionally open-sourced major parts of Claude’s architecture well ahead of schedule. I apologize to the team and to Claude.

@Marie_Haynes Yep EDM cover all the way

@le_kman ahahaha...gave my image to Gemini Lyria3 and asked it to make music out of it.

Not seeing even one site with obvious core update like movement yet, but then again GSC is only reporting up till the 29th, which is two days into the March core update. It's so much easier for me to check now. Vibe coded a tool to show me patterns across hundreds of sites I monitor.

Heads-up: there is something bizarre going on with Google Search Console data right now. Similar to the changes that came to light after the disabling of &num=100, impressions are again skyrocketing for specific surfaces on desktop. For example, using the 'merchant listings' search appearance filter, which has historically been independent of the influence of rank trackers (because of an impression being recorded only after a grid result is selected), is now a mess. Across several large-scale eCommerce sites, CTR data is no longer accurate for desktop, with there now being many queries appearing that are clearly related to tools, with significant increases in impressions from this past week in particular. But it doesn't just seem to be rank trackers that are causing the inaccurate data; I'm also seeing the consistent appearance of a "product" query for several unrelated businesses. That query does not trigger a merchant listing-related result, so it makes no sense. As a side note, I'm also seeing similar within Google Images when using the 'Images' filter, where desktop impressions are again very out of whack. Because tracking standard organic listings is now more complex for scrapers, it seems like they're now looking to more untapped surfaces, such as product grids and Google Image results for insights in tools. As I mentioned, the scraper services only seem to be part of the issue. I'm seeing far more data inaccuracies than I have in the past for these two surfaces. It would be great if the Google Search Console team could look into this, as it now makes our datasets less useful, especially within the eCommerce segment.

The part nobody's thinking about: this loop has a termination point. Right now, a lab lands a breakthrough, uses it internally for 3 to 6 months, ships a wave of products, then releases the model publicly. The public gets the breakthrough. The gap closes. Cycle repeats. But once the internal model is good enough to meaningfully accelerate training the next model, the public release becomes a farewell letter. You're getting the version they've already moved past. The version that already built its replacement. Watch Anthropic right now. 74 releases in 52 days. Claude Code, Cowork, Computer Use, Channels, Security, Opus 4.6, Sonnet 4.6. Cowork was built with Claude Code in 10 days. Their own engineers ship 60 to 100 internal releases per day. Then on Thursday a misconfigured CMS leaked a draft announcement for "Claude Mythos," a new model tier above Opus that Anthropic confirmed is real and called "a step change." They've been running on something the rest of us haven't seen yet. That's why the shipping velocity looked impossible. The gap between internal capability and public capability has been roughly one model generation. That gap is about to become permanent. Because every month the better model stays internal, it compounds. It writes better training code. It finds better architectures. It debugs its own evals. The "flurry of releases" Kevin describes is the last phase where outsiders can even track what's happening. After that, the velocity becomes unexplainable without knowing what's underneath it.

Fun fact: the site: search operator only works on Google and Bing, but not Exa. ChatGPT uses site: search operators in fan-out queries. 😅

@lilyraynyc I would not like my clone 😒

Has anyone made an AI agent I can use to clone myself to be able to attend all the dozens of product demos I am pitched weekly?

Nice one .. I love sci-fi and seo so easy read

Why Google’s New “Google-Agent” Is The Biggest Mindset Shift In SEO History. via @marie_haynes: hubs.li/Q048Cfn20 #Google #SEO

This week, Google announced a huge breakthrough, called TurboQuant. This is a really big deal. TurboQuant makes it significantly faster to assemble and query a vector space. Their abstract says that it reduces the time needed to build and index for vector search to "virtually zero." This has profound implications for Search. Although Google released their blog post on TurboQuant this week, the research paper behind it was published a year ago, which gave them plenty of time to implement the technique into their core systems. Perhaps the March core update will introduce us to TurboQuant. If Google has started using TurboQuant, then I expect we will see some significant changes. One is good for website owners. And one is not. 1) Google should get better at truly surfacing great, original and insightful content that meets the searcher's needs. If you're focusing on creating awesome stuff rather than commodity content, this update should be good for you. but also, 2) TurboQuant will allow for more AI Overviews. Google will be able to assemble the content needed to produce an AI Overview instantly. They'll have more confidence in their ability to answer even super complicated queries. I wrote a blog post breaking it down: mariehaynes.com/turboquant-has…

@CyrusShepard Well.. EU and UK are stricter than the us JD

Google helped grow the internet until they became the internet. They were declared an illegal monopoly, but a judge declined to break them up. Now with every new update, Google takes a little bit of traffic away from the open web and gives it to themselves. They will continue this until they can't. Shareholders profit, and websites wither on the vine.

Outstanding research from @TheCoolestCool on Reddit's absolutely wild domination of Google's B2B results: foundationinc.co/lab/reddit-b2b… Just look at these numbers 🫪🧐🤯

@CyrusShepard My guess.. cost for generating this is high, returns are low

Friends, Google currently has me in a test where it shows no AI Overviews It's so nice

@Marie_Haynes @LizMicik I think there's a bigger money grab in authenticating human creations.. ibl would not underestimate our need for human connections even with products

Soon, your agents will negotiate with my agents, and eventually those who understand how agents work will make a lot of money. Full chat with @LizMicik on AI agent protocols marketers need to know is here: youtu.be/i_sAhytJFg8

Last week's SEO news, guides, analysis in 13 min :) youtu.be/3bLboYjRcxs

SO MANY people are getting misled by "AI citation" research studies, and the concept of what it means to be "cited" by an AI response. I'm no expert, but the people who answered this question are: docs.google.com/document/d/109… Thanks to @lucatag2015 for publishing this piece.