Ahrefs Digest

I’m not intentionally curating against a weekly theme, but somehow the posts I’ve read and liked this week are all about AI content.

(By the way, if you prefer me to curate with a weekly theme in mind, let me know. And tell me why too!)

Basically, throw out the old content marketing playbook, embrace AI content (but for the robots), and reserve the human content for well, the humans. And oh, try not to let AI-isms into your human content. 

Trust me, you will, at some point. It’s like cooking, you know. Like it or not, if you’re not looking at the labels carefully, you’ll unintentionally let some ultra-processed foods into your cooking. Even if you’re free-range, organic, UPF-free, or whatever the latest food buzzword is. 

Maybe the predictions were right. Maybe human-written content will become the premium content people pay for, just like how we pay for “hand-cooked chips” and “handcrafted jewelry” today. AI content will be reserved solely for the robots, and for those of us who are not fortunate enough to get access to human content. 

What do you think? Do you agree or do you think something else? Hit ‘reply’ and let me know! 

Meme of the week

 

—

Quick search marketing news

  • OpenAI, Google, Microsoft and Anthropic say they’ll honour one robots.txt line to opt out of AI training.It blocks training only, so bots can still crawl your pages and use them in AI answers.
  • Google says they may keep “important” pages in its index for a while even when they return a 4xx error. They show up with no snippet or favicon, though, so stray error codes still cost you.
  • Google launches a platform for developers to bring their apps into both Search and Gemini. More tasks will get done inside Google without anyone visiting a website.
—

Best of the week

Headless Content Marketing: A New (Old) Strategy for Search: More big brands seem to be publishing pages written only for AI to read. They’re terse, packed with facts and brand mentions, updated constantly, and usually AI-generated. Should we or can we blame them? We’ve always responded to the incentives Google creates. This is just one example of many. We tried to make SEO content work both ways, for humans and for robots, but maybe the future is bifurcated: make content for only humans, and make content for only robots.

Content Marketing As We Knew It Is Over: Nobody’s clicking, nobody’s reading, nobody fills forms anymore. HubSpot’s SVP Agentic GTM & Systems, Kieran Flanagan, thinks there’s a new playbook, with more focus on creators, originality, and entertainment. In all honesty, I think we were always trending towards this direction. AI just made it faster. Because imo, it’s just human nature. We love learning from other people and hate corporates. And in the past, YouTube and blogs were that. But then came “professionalism” and everything became over-produced and corporate. So, short-form videos and creators are simply the newest medium that allows us to access people again. (Well, until it changes again.)

Claudish Is Sabotaging Your Writing: Joe Lazer argues that reading AI-generated content primes us to copy its sentence structures and word choices, which linguists call structural and lexical priming. Yes, my friends. Even if you’re not using AI to make content, reading too much of it (whether we like it or not, it’s here) will still cause us to use AI-isms in our writing. Without realizing it. The horrors!

How Does the LinkedIn Algorithm Work in 2026? 116,000+ Posts Analysed: Flypost tested the “LinkedIn best practices” against 116,000+ posts from 2,400+ B2B creators, comparing every creator against their own posts. Three things clearly work: personal stories, images and writing in the first person. Despite all the slop on LinkedIn, and all the reporting of “this is AI slop”, looks like there’s still room for humans…?

—

Latest from our podcast

 

This week’s podcast guest is Eli Schwartz, author of Product-Led SEO and a solo consultant for clients like Tinder, Coinbase and LinkedIn. 

Eli is here to talk about what SEO consulting pays, and why. His math for a $1M practice starts with a $10,000-a-month floor, and at the top of the range, he knows solo consultants earning more than $100,000 a month from a single client. They're charging for value, not hours.

Eli explains why Google isn't nervous about ChatGPT anymore, why getting recommended by AI comes down to building a real business, and why a prompt-tracking dashboard that can't show revenue is worth little. You'll come away knowing what to charge, how to land the clients who can pay it, and what to measure when the CEO asks what AI visibility is worth.

Watch the full episode on YouTube.

 

See you next week,
Si Quan

P.S. Liked this week’s Digest? Forward it to a friend or tell them to subscribe here.

Honest verdict on this one?

🔥  Banger 🙂  Fine 😴  Snooze