Hey there,

Welcome back to another Don’t Feed The Algorithm, your weekly dispatch from Averi — where we explore the intersections of marketing, AI and the future of work.

The playbooks merged this week. Nobody announced it. Nobody published a press release.

But the data is now unambiguous… the content that wins in Google's core update and the content that gets cited by AI search platforms are the same content.

Original data. Named experts. Verifiable claims. Front-loaded answers. Fresh updates. Information Gain.

The signals Google rewarded in its March core update are the same signals the AirOps fan-out study showed ChatGPT selects for. The correlation between AI usage and ranking penalties is 0.011.

Google doesn't care if AI helped you write. It cares whether you added something the internet didn't already have.

Meanwhile, OpenAI killed Sora — a product that was generating spectacle without sustainable value — and redirected every GPU toward enterprise AI. Anthropic's MCP hit 97 million installs. Google expanded AI search to 200+ countries. The infrastructure is consolidating around quality, not volume.

This issue is built around that convergence. The Information Gain Sprint gives you a 90-minute playbook for adding original data to your most important pages. What The Bots Are Reading breaks down why HubSpot lost 70-80% of its organic traffic while sites with original research gained 22%. Feed the Engine — a new section starting this week — teaches one content engineering concept per issue, beginning with schema markup. And Nike reminded everyone that athletes over algorithms isn't just a marketing slogan. It's the winning strategy.

Let's build.

— ZC & The Averi Team

Schema markup is invisible code on your page that tells search engines and AI systems what your content is — not just what it says.

It's the difference between a bot reading your page and guessing you're an expert, and your page explicitly declaring: this is an article, written by this person, with these credentials, published on this date, updated on this date, answering this question.

Three types matter most right now.

Article schema tells Google and AI crawlers who wrote the piece, when, and what organization published it — feeding directly into E-E-A-T signals.

FAQ schema marks up your questions and answers so they can be pulled into featured snippets and AI-generated responses without the system having to guess which part of your page is the answer.

Organization schema establishes your brand as a known entity across the web, which strengthens every other signal attached to your domain.

Pages with valid structured data see a 73% higher AI selection rate. Most content platforms don't add it automatically. Check yours this week — if your blog posts don't have Article schema with an author field, you're leaving citation share on the table.

Next week: Front-Loading Answers — why the first 30% of your page captures 44% of AI citations.

Here's the problem we kept running into: you publish a piece of content, it looks great, it reads well — and then it just sits there. Doesn't rank. Doesn't get cited by ChatGPT. Doesn't show up in Perplexity. You're left wondering what went wrong, and the answer is usually buried across six different tools and a dozen blog posts about "best practices."

So we built a scoring system that tells you before you publish whether your content is optimized for both Google and AI search — in one number.

Every piece of content in Averi now gets an SEO + GEO composite score (0-100). SEO is weighted at 55%. GEO is weighted at 45%. The score updates in real time as you write.

Here's what it actually checks:

→ SEO signals — keyword coverage, content structure, heading hierarchy, readability, meta title and description, internal links, image alt text. The fundamentals that still drive Google rankings.

→ GEO signals — citation-worthiness (do you have statistics with attribution? original claims? expert quotes?), extractability (are your answers front-loaded in 40-60 word blocks that LLMs can pull cleanly?), entity density (named people, companies, frameworks), E-E-A-T markers, and whether your site even lets AI crawlers in.

→ Actionable recommendations — every check that scores below its max generates a specific fix, prioritized by how much it would improve your composite score.

Not: "improve your content quality."

More like: "Add citations from different authority types (government, academic, industry)"

Why this matters right now: every piece of data in this newsletter says the same thing. Being found isn't enough — you have to be selected. ChatGPT throws away 85% of what it retrieves. Google's AI Overviews are driving impressions up but traffic down. The content that wins is structured for extraction, loaded with citable data, and optimized for both the algorithm and the AI. That's exactly what the scoring system measures.

No other content platform scores for GEO. That's the gap we're closing.

— Averi Team

Every section of this newsletter exists because the rules of content marketing are being rewritten in real time. Fan-out queries. AI citation surfaces. GEO. LinkedIn as a citation source. Content scoring. None of this was in anyone's playbook two years ago.

We built Averi Academy to help you learn all of it.

It's a free certification program designed to turn content marketers into AI-powered content marketers — the people who understand how to create content that ranks on Google and gets cited by ChatGPT, Perplexity, and Copilot. You'll learn the frameworks, workflows, and strategies we write about in DFTA every week, structured into lessons you can actually execute on.

Finish the program, earn a certificate you can add straight to your LinkedIn profile.

Whether you're a founder running marketing solo, a content marketer leveling up your skillset, or someone eyeing one of those roles in Don't Let AI Replace You — this is built for you.

👉 Become a Certified AI Content Engineer →

AI search engines now handle 60% of B2B product discovery. Google organic CTR has dropped 61% on queries with AI Overviews. And most marketing teams are still running the same playbook from 2023.

We wanted to know: what happens when an entire industry is forced to adapt at once?

So we ran a swarm intelligence simulation (12 AI agents representing CMOs, VPs of Marketing, SEO managers, content leads, and demand gen specialists) and let them debate, strategize, and compete across 15 rounds of simulated market disruption.

The results weren't subtle.

Google and the AI platforms just agreed on what good content looks like. They didn't coordinate. They converged.

The March 2026 core update is completing this week — and a JetDigitalPro study of 600,000 pages reveals a pattern that matters far beyond Google rankings… the content winning in the core update is the same content getting cited by AI search platforms.

Here's the data.

AI content farms lost 60-80% of organic traffic. Affiliate sites took a 71% negative hit — the hardest-hit category. Content not updated within 90 days lost 20-40%. But sites publishing original research gained an average 22% visibility boost. 72% of top-ranking pages now feature detailed author credentials, up from 58% before the update. And the correlation between AI usage and ranking penalties was 0.011 — essentially zero.

Google isn't penalizing AI-assisted content. It's penalizing content that exists only to intercept search traffic without adding anything new.

The key signal is Information Gain. Google has patented the concept and discussed it in research papers for years, but this is the first update where it's showing up with real teeth. Information Gain measures how much genuinely new knowledge your page contributes compared to what already ranks. Pages that reword the top 10 results without adding fresh data, perspectives, or analysis are dropping. Pages with proprietary data, first-hand case studies, and expert commentary that can't be found elsewhere are climbing.

Here's why this matters for AI citation too: the same qualities that Information Gain rewards — original data, named experts, verifiable claims, structured answers — are exactly what AirOps' fan-out study showed ChatGPT selects for when choosing which 15% of retrieved pages to cite.

And it's what Ahrefs found in the AI Overview decoupling: only 38% of citations come from top-10 pages now, because AI systems are selecting for content quality and specificity, not just rank position.

The convergence is structural, not coincidental. Google is using Gemini 4.0's Semantic Filter in this update. AI Overviews draw on the same quality signals as organic rankings. ChatGPT's retrieval-to-citation pipeline rewards the same attributes — original data, attribution phrases, entity density, front-loaded answers.

The playbook for "get ranked by Google" and the playbook for "get cited by AI" just merged into one playbook.

And there's one more data point that should keep founders up at night: HubSpot reportedly lost 70-80% of its organic traffic across the 2025-2026 update cycle. The former gold standard of content marketing at scale — undone by years of publishing broad top-of-funnel content with weak connections to its core product. If HubSpot's content strategy can't survive this environment, no one's can. Volume doesn't compound anymore. Quality does.

What this means for startups: Stop treating SEO and GEO as separate disciplines. The content that ranks in Google and the content that gets cited by AI are now evaluated on the same signals: original data, named expertise, information gain, and structured extractability. One strategy. One standard. Build for both or lose to both.

📊 By the numbers:

We all know the industry loves to pretend it’s got it all figured out.

But here?

We tell the truth.

Every week, we’re asking you to share how things really work behind the scenes. Because getting better starts with getting honest.

Don’t worry, it’s a safe space.

Drop your vote.

Nike — "So Win"

Nike hadn't run a Super Bowl ad in 27 years. When it came back, it didn't run a product spot. It didn't run a celebrity cameo. It ran a 90-second black-and-white anthem about female athletes being told what they can't do — and doing it anyway.

Caitlin Clark. Sha'Carri Richardson. A'ja Wilson. Jordan Chiles. Sabrina Ionescu. Aryna Sabalenka. Each one shown in the middle of the thing they were told they couldn't do — breaking records, filling stadiums, demanding attention — while rapper Doechii narrated every criticism women athletes hear, building to a three-word answer: "So Win."

Created by Wieden+Kennedy Portland. Directed by Kim Gehrig. Set to Led Zeppelin's "Whole Lotta Love." No product shots. No app downloads. No QR codes. Just athletes and a point of view.

Why we love it: Nike's CMO said the company is getting "back to athletes over algorithms." That single line is the most relevant sentence in marketing right now — the same week Google's core update proved that original, expert-driven content beats algorithmic content at scale, Nike proved the same thing in advertising. The campaign didn't optimize for engagement metrics. It optimized for meaning. And the meaning did the distribution.

The numbers were staggering. 66 million Instagram views in 24 hours — Nike's most-watched video ever on the platform. 28 million views across the follow-up athlete spotlight series. Won the Super Clio for best ad of the game. Nike's logo appeared 819 times across 521 frames during the broadcast — more than any other brand. The "So Win" T-shirt featuring all six athletes sold out immediately. A'ja Wilson's signature shoe, the "A'One Pink Aura," sold out within minutes when it dropped months later.

And here's the strategic depth most coverage missed: Nike didn't run a Super Bowl ad in 2026. They came back for one year — proved they could still own the biggest stage in advertising — and then stepped away again. They chose the moment, executed perfectly, and walked away before it became routine. That's not a media strategy. That's restraint as a competitive advantage.

The algorithm could never understand that the most powerful brand statement Nike could make after 27 years of silence was a 90-second film with no product in it — and that the second most powerful statement was not coming back the following year.

The startup takeaway: When you finally have something worth saying, say it once, say it perfectly, and don't dilute it by saying it again next week.

🔗 Watch the ad

Strategy You Can Steal

The Information Gain Sprint

What it is: A 90-minute sprint to identify your highest-value pages, diagnose what original data they're missing, and add the specific signals that both Google's core update and AI citation systems reward.

Time commitment: 90 minutes, 4 blocks

Block 1: Find Your Vulnerable Pages (20 min)

Open Google Search Console. Filter to the last 28 days. Sort by clicks, descending. These are your highest-traffic pages — the ones where a core update loss hurts the most. Now cross-reference: for each of your top 15 pages, ask two questions. First, does this page contain any data, stat, or claim that exists only here? If the answer is no, it's vulnerable. Second, when was it last updated? Content not touched in 90+ days lost 20-40% in this update. Flag every page that fails either test.

Block 2: Mine Your Own Data (25 min)

You have original data. You're just not publishing it. Open your analytics, your CRM, your product dashboard, your customer success records. Look for: conversion rates from your own funnel, time-to-value metrics from your onboarding, before/after results from customers, engagement benchmarks from your content, pricing comparisons you've run internally, survey results from your audience. Write down 10-15 data points. They don't need to be groundbreaking. They need to be yours. "Our customers see an average 34% increase in organic traffic within 90 days" is more valuable to Google and AI systems than a hundred rewritten paragraphs from someone else's blog post.

Block 3: Add Information Gain to Your Top 5 (30 min)

Pick the five pages from Block 1 that are most vulnerable and most important. For each one, add at least one of the following: a proprietary statistic with clear attribution ("Based on Averi's analysis of 500 content campaigns..."), a named expert quote — even if the expert is you (72% of top-ranking pages now show author credentials), a first-hand case study with specific numbers, a comparison table built from your own testing, or an original framework or methodology that can be cited by name. Structure it for extraction: put the data point in the first 30% of the relevant section. Use attribution language ("According to," "Our data shows," "Based on analysis of"). Include the number, the source, and the context in a single sentence an AI system can pull cleanly.

Block 4: Update and Timestamp (15 min)

For each page you edited, update the "Last Updated" date. Add or update the author bio with verifiable credentials. If you added a new section, make sure the H2 is a question your audience would actually ask — question-based headings still drive both featured snippet selection and AI extraction. Push the updates live. Then open Bing Webmaster Tools and submit the URLs through IndexNow so AI systems see the fresh content faster. Screenshot your Search Console baseline for these five pages. Check again in 30 days.

Why this works: The March 2026 core update and AI citation systems are now rewarding the same signals. Original data. Named expertise. Verifiable claims. Fresh content. The correlation between AI usage and ranking penalties is 0.011 — Google doesn't care if you use AI to write. It cares whether you added something the internet didn't already have. This sprint makes sure your most important pages pass that test for both Google and the bots.

Know of a role that should be here? Reply to this email and we'll feature it next week.

Did You Know? —  86.5% of top-ranking pages use AI assistance. The correlation between AI usage and ranking penalties is 0.011.

Google doesn't care if AI helped you write. It cares whether you added something that wasn't already there.

Til next time,