🐊 Editor's Note

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.

Citation.

That's the word this week. Not ranking. Not click-through. Citation.

The Seer Interactive study that dropped this month is the clearest data we've seen on what's actually happening in search: organic CTR is down 61% when an AI Overview appears, and 43% of those sessions end in zero clicks. But brands cited inside the Overview earn 35% more organic and 91% more paid clicks than brands that aren't.

The math has flipped. Being named inside the AI summary is now a more valuable asset than ranking below it.

Pair that with the release cadence this week — Claude Opus 4.7 shipping with high-res vision and task budgets, Meta's Muse Spark launching across every surface they own, Perplexity's revenue jumping 50% on agentic workflows, OpenAI launching a vertical life sciences model and a $100/month Pro tier — and the story isn't "AI is coming." AI is already how your customers are finding, evaluating, and deciding. The distribution layer of the internet moved, and most startup content strategies are still targeting the old one.

This week's What The Bots Are Reading breaks down the Seer data in full, the Strategy You Can Steal is a 90-minute citation audit you can run before Thursday, and The Algorithm Could Never features The Halal Guys' pitch-perfect April Fools campaign that reminded us why origin stories still outperform ad spend.

The good news, if you're a startup founder reading this at 11pm between Stripe dashboards and a sales email you're not sure you should send… the citation economy rewards the things you already have. A clear point of view. Original data. A specific brand voice. Content structured around what your customers actually ask. The bots aren't waiting for polish. They're waiting for clarity.

Let's build.

— ZC & The Averi Team

Was this newsletter forwarded to you? Sign up to get it in your inbox.

🔥 In Case You Missed It…

  • Claude Opus 4.7 shipped with high-res image support and task budgets — Anthropic released Claude Opus 4.7 on April 16, a meaningful step up from 4.6 on agentic coding and long-horizon reasoning. It's the first Claude model with high-resolution image support (2576px / 3.75MP) and introduces "task budgets" that give the model a countdown of tokens to finish work gracefully. Pricing stays flat at $5 per million input tokens and $25 per million output tokens. Anthropic also admitted Opus 4.7 trails its unreleased cyber-focused Claude Mythos model in some benchmarks, which is unusual honesty for a launch post. → Anthropic · CNBC

  • OpenAI launched GPT-Rosalind for life sciences research — OpenAI introduced a frontier reasoning model purpose-built for biology, drug discovery, and translational medicine, available as a research preview in ChatGPT, Codex, and the API through a trusted access program. Launch partners include Amgen, Moderna, the Allen Institute, and Thermo Fisher. The vertical AI push signals OpenAI sees more enterprise revenue in domain-specific reasoning than in general-purpose chat. Business customers went from 20% of revenue in 2024 to 40% today, and OpenAI expects it to hit 50% by year-end. → OpenAI · Fierce Biotech

  • Meta debuted Muse Spark after a $14B infrastructure overhaul — Meta launched Muse Spark on April 8, the first model from Meta Superintelligence Labs under new AI chief Alexandr Wang. The natively multimodal reasoning model is described as "small and fast by design" with a Contemplating mode for harder queries, and it will roll out across Facebook, Instagram, WhatsApp, Messenger, and Ray-Ban Meta AI glasses in the coming weeks. Meta is spending $115B-$135B on AI capex in 2026, almost double last year. → Meta · TechCrunch

  • Brands cited in AI Overviews earn 35% more organic clicks, 91% more paid — New data from Seer Interactive analyzing 3,119 queries across 42 organizations confirmed what GEO practitioners have been saying for a year: citation is the new ranking. Organic CTR drops 61% when an AI Overview appears, and 43% of AI Overview sessions end in zero clicks. But brands cited inside the Overview earn 35% more organic and 91% more paid clicks than brands that aren't. The correlation between branded web mentions and AI citations is 0.664; for backlinks, it's 0.218. → PragoMedia · Seer Interactive

  • Google confirmed the March 2026 core update is complete — Google confirmed the March 2026 core update finished rolling out on April 8 after a 12-day deployment that began March 27. The parallel March spam update wrapped in under 20 hours, the shortest confirmed spam update in dashboard history. Google guidance this cycle was less about punishment and more about re-ranking based on intent match, freshness, and experience signals. Sites with weak differentiation or misaligned intent saw the sharpest declines, while topical authorities mostly held or grew. → Search Engine Land · Search Engine Journal

  • Perplexity revenue jumped 50% after pivot to AI agents — Perplexity's shift from search to agentic workflows has paid off, with estimated ARR topping $450M in March, up 50% month over month. This week the company shipped Skills, a Model Council feature, Voice Mode, and a GPT-5.3-Codex coding subagent inside Perplexity Computer, plus added GPT-5.4 for Pro and Max subscribers. Deep Research now generates presentations, spreadsheets, dashboards, and websites in-product. A new tax agent also shipped, extending the agentic surface area into compliance work. → PYMNTS · Perplexity Changelog

  • OpenAI launched ChatGPT Pro at $100/month, pivoted on shopping — OpenAI rolled out a new $100/month ChatGPT Pro tier with 5x the Codex limits of the $20 Plus tier, plus shipped over 90 curated plugins for skills, app integrations, and MCP servers. Workspace analytics replaced user-level analytics for Business admins. The company also scrapped Instant Checkout inside ChatGPT, saying the first version "did not offer the level of flexibility we aspire to provide." Merchants now use their own checkout experiences while OpenAI focuses on product discovery. → CNBC · TechCrunch

  • HubSpot priced Breeze AI agents per outcome, not per seat — HubSpot's shift to outcome-based pricing for its Breeze AI agents — $0.50 per resolved conversation and $1 per qualified lead — is a direct response to enterprise buyers tired of paying for AI potential. It's the clearest signal yet that the SaaS pricing model is fracturing under agentic AI. Meanwhile, HubSpot's flagship INBOUND conference has officially been renamed UNBOUND for its September 16-18 return in Boston, reflecting a broader category rebrand around AI-era growth. → Marketing Trends April 20 · UNBOUND 2026

👾 Feed The Engine

This week: Topical Authority

Google doesn't evaluate pages in isolation anymore. It evaluates how much your entire domain knows about a topic.

That's topical authority — and it's now the single strongest predictor of AI citation, with a 0.41 correlation. Domain authority, by comparison, explains less than 4% of variance.

Here's how it works. If you publish one article on "content marketing for startups," Google sees a page. If you publish 15 interconnected pieces — covering strategy, execution, measurement, tools, case studies, mistakes, and trends — Google sees an expert. The depth and breadth of your coverage signals that you don't just know the answer to one question. You know the entire subject.

The compounding effect is real. A narrow site covering one domain deeply outperforms a broad site covering ten topics at surface level. Google's March core update made this explicit… sites with comprehensive topical coverage gained visibility while sites that published widely but shallowly lost it. And AI citation systems follow the same pattern — ChatGPT and Perplexity are more likely to cite a source they've seen answer multiple related questions consistently than one they've seen answer a single question once.

You can't build topical authority in a week. But you can start this week.

Pick your single most important topic. Map every question your audience asks about it. Check how many of those questions you've published answers for. The gaps are your content queue. Fill them — not all at once, but consistently — and the compounding starts.

Next week: Entity Density — what "15+ named entities = 4.8x more citations" actually means in practice.

⚙️ Straight Outta Dev

Your Strategy Map is now yours to edit — add, delete, and rebuild pillars on your terms.

When we launched Strategy Map, Averi generated your content pillars based on your Brand Core — your ICP, your positioning, your competitive landscape.

It was a strong starting point. But starting points aren't enough when your strategy evolves, your market shifts, or you realize a pillar that looked good on paper isn't driving results.

Now you have full control.

Three new capabilities are live inside Strategy Map:

Add your own pillars. See a topic your audience cares about that Averi didn't surface? Add it yourself. Name it, scope it, and Averi will generate focus areas and topics underneath it — fitting your custom pillar into the same strategic framework as everything else in your map.

Delete pillars that aren't working. Not every pillar earns its place. If a pillar isn't driving traffic, isn't relevant to your ICP, or was never right to begin with — remove it. No guilt. Your Strategy Map should reflect what you're actually building authority on, not what sounded good three months ago.

Regenerate pillars with AI. Want to keep the topic but rethink the approach? Hit regenerate and Averi rebuilds the pillar — new focus areas, new topic suggestions, new angles — using everything it knows about your brand, your audience, and what's performing in your space right now.

View your entire map at once. Until now, you navigated your Strategy Map one pillar at a time. Now you can expand everything — every pillar, every focus area, every topic, every sub-topic — in a single view. See the full architecture of your content strategy laid out in front of you. Spot gaps. Find overlaps. Understand how your pillars connect. The bird's-eye view turns strategic planning from clicking through layers into seeing the whole system at once.

This matters more than it sounds. Your Strategy Map is the foundation that drives your Content Queue, your scoring, and now your Priority Next Actions. Every pillar you add, remove, or rebuild cascades through the entire engine. A sharper map means sharper recommendations, sharper content, and sharper results.

This is what we mean by guided, not autopilot. Averi builds the strategic framework. You shape it. The engine runs on both.

— Averi Team

📚 New: Welcome To Averi Academy

The data in this week's issue points at one difficult reality. The skills that made a great content marketer in 2022 — keyword research, on-page SEO, backlink outreach — are necessary but no longer sufficient.

You need to know how LLMs retrieve, how citations get earned, and how to structure a piece so it's extractable by machines and compelling to humans at the same time.

That's what Averi Academy teaches.

It's free. It's built for operators. And it ends with a Certified AI Content Engineer credential that actually means something when a recruiter pulls up your LinkedIn.

The curriculum covers the new stack end to end — brand core systems, content queue design, GEO-first drafting, AI citation tracking, and the editorial standards that separate extractable content from generic AI slop. Every module is taught by practitioners who've actually grown traffic in the AI search era, not consultants recycling 2023 slides.

If you're hiring for one of the GEO roles we listed above, or applying for one, this is the fastest way to get the vocabulary and the portfolio to move.

📝 The Good Sh*t

The Authenticity Premium: Why Human-in-the-Loop Content Outperforms Pure AI by 4x

CNN called it in December 2025: 2026 will be the year of "100% human" marketing.

iHeartMedia found that 90% of its listeners want their media created by humans — even the ones who use AI tools themselves.

Apple TV's "Pluribus" from Vince Gilligan ran credits that read "This show was made by humans."

The Tyee adopted a no-AI journalism policy. Pinterest's AI embrace is alienating its most dedicated users. New York subway ads for AI products get vandalized with "AI is not your friend."

Merriam-Webster's 2025 Word of the Year? "Slop."

If you're a startup founder reading this and thinking "great, so I should stop using AI for content," hold on.

That's not what the data actually says.

The data says something more nuanced and more useful:

The brands winning in 2026 aren't the ones avoiding AI. They're the ones using AI as infrastructure and humans as voice.

More From The Averi Journals

  1. Agentic Marketing for Startups: What It Means (And What to Ignore)Read online

  2. We're Building for the Founders Nobody's Building For Read online

  3. Best AI SEO Tools for Small Marketing Teams (2026) Read online

🤖 What The Bots Are Reading

The click is dead. The citation is the new asset.

Seer Interactive dropped the most important piece of GEO research of the year this month, and if you haven't internalized it yet, your 2026 content strategy is already wrong.

Between June 2024 and September 2025, Seer analyzed 3,119 informational queries across 42 organizations.

They tracked 25.1 million organic impressions and 1.1 million paid impressions. The scale matters because the conclusions are big enough to kill a decade of SEO orthodoxy.

Here's what they found.

When an AI Overview appears on a search result page, organic click-through rate on the top result drops 61%. Paid CTR drops 68%. About 43% of AI Overview sessions end in zero clicks — the user reads the summary and leaves. This tracks with Ahrefs' parallel study from February 2026, which put the drop at 58% on average.

The old playbook said: rank #1, get the click. That playbook assumed the click existed. It doesn't anymore for huge portions of informational search.

But here's the twist that changes the math entirely. Brands cited inside the AI Overview earn 35% more organic clicks and 91% more paid clicks than brands not cited — on the same query. Being inside the summary is now worth more than ranking below it. Citation has replaced ranking as the visibility asset.

The research that makes this actionable is the correlation work. The correlation between branded web mentions and AI Overview citation is 0.664. The correlation between traditional backlinks and AI citation is 0.218. In the old world, backlinks were the strongest signal. In the new world, brand mentions are three times stronger.

This flips a lot of agency playbooks on their head. Link building, guest posting, and PR that chased dofollow links now matters less than unlinked brand mentions across authoritative sites, Reddit threads, YouTube descriptions, and industry newsletters. The signals the bots care about are the signals humans were already producing organically — talking about brands in context.

Parallel research from Omniscient Digital, which analyzed 23,000+ AI citations, backs this up. Listicles (21.9%) and articles (16.7%) dominate citations in AI Mode, ChatGPT, and Perplexity. And 44.2% of all LLM citations come from the first 30% of an article — meaning the intro is doing most of the heavy lifting. Add statistical facts and you get a +22% citation lift. Add direct quotations and it's +37%.

What this means for startups: You already do not have budget for a massive link-building operation. Good news: that's not where the visibility is anymore. The startups that win the citation economy are the ones who publish original data, stake specific positions, get mentioned (linked or not) in industry conversations, and structure content for extraction. Front-load your facts. Name your sources. Earn the quote pull.

The ranking race got harder and less rewarding. The citation race just started, and the early signals favor the specific, the data-backed, and the brand-forward. That's a race startups can actually win.

📊 By the numbers

  • 61% — drop in organic CTR when an AI Overview appears (Seer Interactive)

  • 58% — average click reduction found by Ahrefs in February 2026 (Ahrefs)

  • 43% — share of AI Overview sessions that end in zero clicks (Seer)

  • 35% — more organic clicks for brands cited inside AI Overviews

  • 91% — more paid clicks for brands cited inside AI Overviews

  • 0.664 vs. 0.218 — correlation of brand mentions vs. backlinks to AI citations

  • 44.2% — share of LLM citations pulled from the first 30% of an article (Omniscient Digital)

  • +37% — citation lift when content includes direct quotations

🕵 Marketer’s Anonymous

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.

Whichever you picked — do you know what it's actually costing you?

Founder hours. Freelancer invoices. Agency retainers. Revision rounds. The campaign that took three weeks instead of three days. The blog post that cost $800 and got 47 pageviews.

We built a calculator that shows you the real cost of your current content operation — and what it looks like when a content engine handles strategy, drafting, publishing, and analytics in one workflow.

See what you could save 💸


📺 The Algorithm Could Never

The Halal Guys — "53rd & Sixth" Fine Dining Concept

On April 1, The Halal Guys announced the opening of 53rd & Sixth, a tasting-menu fine dining concept at the exact Manhattan intersection where the brand started as a street cart in 1990. The mock campaign promised gold-plated white sauce service, 24-karat falafel, Halal-certified Wagyu platters, and tableside pairings from a dedicated "sauce sommelier." Marketing materials were shot in the visual language of Michelin hospitality — crisp linens, black-tie service, soft jazz — but with a chicken-over-rice platter at the center of every hero shot.

The joke wasn't subtle. That's why it worked.

Rather than building a stunt around the brand's current menu, The Halal Guys built a stunt around its origin story. Every detail of 53rd & Sixth referenced the cart — the exact address, the white sauce, the red-and-yellow color palette sneaking into the "new" branding. It was nostalgia wrapped in parody, and it read as affectionate rather than cynical.

The campaign spread across Instagram, X, and TikTok organically within hours. Food media picked it up as one of the standout April Fools campaigns of the year, and the brand's original street cart location saw a measurable foot traffic lift the following week as people came to see "where it all started." The company didn't need to book a celebrity or buy ad placement. The concept carried the whole thing.

Why we love it: Most April Fools campaigns try to be clever. This one was confident. The brand trusted its own origin story to do the heavy lifting, and the joke only works if you know the real product is a $10 street cart platter. It rewarded loyal customers without excluding new ones. The best brand humor works like an inside joke that anyone can join.

Results: Coverage in Ad Age, DesignRush, ContentGrip, and Tom's Guide "Best April Fools 2026" roundups. Organic engagement across social channels in the first 48 hours dwarfed typical brand posts, and the original 53rd & Sixth location reported elevated foot traffic the following weekend. Multiple food media outlets ranked it among the top five April Fools campaigns of 2026.

The algorithm could never: An algorithm can generate an April Fools product concept. It cannot decide that the funniest version of your brand is the most expensive version of your cheapest product, at the exact address where your founders first swiped a credit card. That's a human deciding which part of the brand story to exaggerate and which to protect.

The startup takeaway: Your origin story isn't just an About page — it's the rawest material you have for earned media.

Watch the campaign:Best April Fools 2026 Roundup

🗓️ Make Friends

  • NTCA Marketing and Sales Conference | Boston, MA | May 3-5, 2026 | → Register

  • B2B Online Chicago | Chicago, IL | May 4-6, 2026 | → Register

  • Gartner Marketing Symposium | London, UK | May 11-12, 2026 | → Register

  • SaaStr Annual | San Mateo, CA | May 12-14, 2026 | → Register

  • DigiMarCon New England | Boston, MA | May 12-13, 2026 | → Register

  • DigiMarCon Great Lakes | Cleveland, OH | May 28-29, 2026 | → Register

SaaStr Annual 2026

The biggest SaaS founder gathering of the year is back at the San Mateo County Events Center. 10,000+ founders, VCs, and executives across 40 acres with 300+ sessions on go-to-market, pricing, AI agents, and scaling through the weird post-ZIRP market. If your 2026 plan includes raising a Series A or B, pitching enterprise, or just figuring out why your competitor's LinkedIn content is suddenly everywhere, this is where those conversations happen.

Date: May 12-14, 2026

Location: San Mateo County Events Center, San Francisco Bay Area, CA

Link: Register

Gartner Marketing Symposium 2026

If you run marketing at a growth-stage company and want to hear how enterprise CMOs are actually restructuring teams around AI agents, GEO, and outcome-based measurement, this is the event. The two-day London gathering pulls from Gartner's research pipeline and tends to surface the early signals that show up in CMO budget decks six months later. Worth the flight if you're a solo marketer punching above your weight and need to know what enterprise is about to start doing.

Date: May 11-12, 2026

Location: London, UK

Link: Register

💡Ask Averi

AI + Human Marketing Tip

“If 44% of LLM citations come from the first 30% of your article, your intro is no longer a warm-up. It's the asset.”

Go to your three highest-traffic blog posts right now and ask… does the first 150 words answer the question directly, include a statistical fact with a cited source, and name your brand as the entity behind the claim?

If not, rewrite them this week.

That's a 30-minute fix with outsized upside. Most content teams buried the answer three scrolls down and called it engagement. The bots don't scroll. Put the answer where they read.

Strategy You Can Steal

The 90-Minute Citation Audit

You already have content. It just isn't structured to be cited. This sprint turns your existing library into citation-ready assets without writing anything new. Block 90 minutes on your calendar, close Slack, and run it.

Block 1: Pick Your 5 (15 minutes)

Pull your analytics and identify five posts that meet these criteria: ranking on page 1 or 2 for a target keyword, still getting at least 100 monthly impressions, published within the last 18 months. These are your citation candidates. They already have Google's trust. They just need to be made extractable. Skip posts ranking below page 2 — the juice isn't worth the squeeze yet.

Block 2: Run the Intro Rewrite (25 minutes)

For each of the five, rewrite the first 150 words to do three jobs: state the user's question in their own voice as an H2 or opening line, give a 40-60 word direct answer that could stand alone in an AI response, include one hyperlinked statistic from an authoritative source (McKinsey, Gartner, HubSpot, or original data). Keep your brand name inside that first block. LLMs cite entities, not paragraphs.

Block 3: Add the Quote Pull and FAQ (25 minutes)

Insert one pullquote from a named expert (internal is fine if you link to their bio) and add a 3-4 question FAQ block at the bottom with schema-ready formatting. Direct quotations lift citation probability by 37%. FAQ schema is still the cleanest path into both Google featured snippets and AI Overview expansions. Use real user language in your FAQ questions — the exact way someone would type it into ChatGPT, not how a content strategist would write it.

Block 4: Track and Mention (25 minutes)

Set up tracking for the five URLs using whatever GEO tool you prefer (Profound, Peec, SE Ranking's AI tracker, or a spreadsheet if you're scrappy). Then spend the last 15 minutes writing one short LinkedIn post, one Reddit comment in a relevant subreddit, or one newsletter blurb that links the updated post with your brand named explicitly. Brand mentions correlate with AI citation at 0.664. Your five-post audit becomes a five-post mention operation.

Why this works

The Seer data shows citation is now worth more than position. The Omniscient data shows intros do 44% of the citation lifting. This sprint targets both levers at once — making your best content extractable and your brand more mentioned. Five posts, 90 minutes, zero new writing. The bots read what's already there. You're just making sure they find your name when they do.

💼 Don't Let AI Replace You

  • Care.com
    Position: Marketing Manager, SEO & GEO – Austin, TX (Hybrid) Build and execute the SEO roadmap and define Care.com's AI search visibility strategy across Google AI Overviews, ChatGPT, and Perplexity. The job listing literally says "be the answer, not just a result." → Apply Now

  • Ping Identity
    Position: SEO & GEO Manager – Remote Drive SEO and generative AI platform strategies, optimize techniques across traditional and AI search, and collaborate across marketing teams. GEO in the job title. → Apply Now

  • Profound
    Position: SEO & AI Search Optimization Specialist (GEO / LLMO) – Remote Full-time GEO role at one of the leading AI citation tracking platforms. If you want to be at the center of the data that's reshaping how brands show up in AI answers, this is it. → Apply Now

  • Daydream
    Position: Multiple roles (just raised $15M Series A) – Remote The AI-native SEO startup that just closed a $15M round is hiring across the board. If you want to build the next generation of organic search infrastructure from the inside, this is the moment. → Apply Now

  • AllTrails
    Position: Editorial & SEO Content Strategy Intern – Remote World's most popular outdoor platform. Learn content strategy and SEO at a brand with massive organic reach and a product people actually love. Great entry point. → Apply Now

  • BetterHelp Position: Editorial Strategist – Remote Oversee BetterHelp's editorial roadmap, enhance mental health content quality and SEO visibility. Tools listed include AirOps, Profound, and generative search tools — the GEO stack is built in. → Apply Now

  • Anthropic
    Position: SEO Lead – San Francisco, CA Build and scale Anthropic's organic search and AI discoverability program from the ground up. AEO/GEO listed as a preferred skill. → Apply Now

  • Pfizer
    Position: Manager, SEO and AI Discoverability – New York, NY (Hybrid) First-of-its-kind role inside Pfizer's new Content Engine Center of Excellence. SEO + GEO execution across the full web portfolio. $106K-$171K. → Apply Now

  • Pfizer
    Position: Manager, Content Engine – New York, NY (Hybrid) The other half of Pfizer's new SEO/GEO org. End-to-end content production, editorial governance, and quality control for search-driven content at pharma scale. → Apply Now

  • Bitly
    Position: Content Marketing Manager – Remote Own Bitly's content strategy end-to-end: thought leadership, data-driven reports, customer stories, video. AI tools expected. → Apply Now

  • Retell AI
    Position: Content Marketer – Remote SEO-optimized articles, product updates, landing pages, and social for a voice AI startup. Tight product collaboration. → Apply Now

  • EliseAI
    Position: Product Marketing Manager – New York, NY Own positioning, messaging, go-to-market, and sales enablement for an AI-powered housing platform. → Apply Now

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

Did You Know? —  Branded web mentions correlate with AI Overview citations at 0.664, while traditional backlinks correlate at just 0.218 — meaning unlinked brand mentions are now roughly three times more predictive of AI visibility than backlinks.

The citation economy rewards being talked about, not just linked to.

Til next time,

DFTA

What did you think of this issue?

Login or Subscribe to participate

If this hits home, we’d love it if you shared it. Send it to a friend, a teammate, or that one person who’s always drowning in a Google Doc.

📝 Have an idea for an article or column? Pitch us

🔵 Want DFTA on LinkedIn? Subscribe

Reply

Avatar

or to participate

Keep Reading