The AI Search Manifesto: Why Every Company Needs to Become a Publisher or Disappear From the Conversation

We have been building toward this piece since we started State of Brand. 

Every article we have published on AI search visibility, on LinkedIn as a citation source, on thought leadership economics, on the death of gated whitepapers, on Zoom assembling a SWAT team for ChatGPT, on Ramp building a media brand inside a fintech company, on owned media infrastructure has been circling the same argument. We kept finding more data. The data kept saying the same thing.

So here it is. The full argument. The manifesto.

The way people discover companies has fundamentally changed. AI is now the front door for a majority of B2B buyers. The brands that show up inside those AI answers are the ones that publish original, expert-driven, structured content at a pace and quality that AI systems treat as authoritative. The brands that do not publish get described by whatever scraps the model can find, or they get left out entirely.

Every company needs to become a publisher. Not eventually. Now.

Part One: The Old Discovery Model Is Broken

For 20 years, discovery worked like this. A buyer had a question. They typed it into Google. They clicked on a link. They landed on your website. You had their attention.

That model is breaking apart in real time.

Gartner predicted in early 2024 that traditional search engine volume would drop 25% by 2026. That prediction appears to be tracking. Google referral traffic to publishers dropped 33% globally in the year to November 2025, according to Chartbeat data published in the Reuters Institute's journalism trends report. For smaller publishers, referral traffic from traditional search declined 60% over two years.

When Google's AI Overviews appear, which they now do on 18% of all searches and 57% of long-tail queries, 43% of those searches end without a single click to an external website. When Google's full AI Mode is active, that number hits 93%.

Ninety-three percent zero-click.

At the same time, the trade publications that B2B companies relied on for decades to reach their buyers are shuttering, consolidating, or cutting staff at historic rates. Publishers have reported losing 20%, 30%, and in some cases 90% of their traffic and revenue over the past year. The Reuters Institute projects more closures ahead.

The old model was: rank on Google, get clicks, capture leads. That model is not dead. But it is shrinking, and the companies still treating it as their primary discovery channel are building on ground that is visibly eroding beneath them.

Part Two: AI Is the New Front Door and the Data Is Not Close

Here is the number that changed everything for us.

G2's latest research, published in April 2026 and based on a survey of more than 1,000 B2B software buyers, found that 51% now start their research with an AI chatbot more often than with Google. That number was 29% when G2 published its 2025 Buyer Behavior Report. It nearly doubled in less than a year.

Seventy-one percent rely on AI chatbots at some point during their buying process. And 87% of B2B software buyers say AI chatbots are changing the way they research new products and services.

ChatGPT alone now fields queries from over 900 million weekly users, according to data compiled by Backlinko. AI chatbot sessions have been doubling annually, reaching 1.2 billion monthly visits by 2026.

And the visitors who come through AI convert at rates that should make every performance marketer pay attention. Semrush found that the average LLM visitor converts at 4.4x the rate of the average organic search visitor. Ahrefs ran the numbers on its own site: 0.5% of visitors came from AI, but those visitors drove 12.1% of signups. A 23x conversion rate multiplier. Opollo's 2026 benchmark report found AI visitors converting at 14.2% versus 2.8% for Google organic. In one firm's data, AI accounted for just 4% of sessions but 19% of the qualified inbound pipeline.

The volume is still small today. The value per visitor is enormous. And the volume is doubling every year. This is early Google in 1999. The companies that figure out how to show up in these answers now are building a lead that compounds.

Part Three: What AI Cites Is Not What Google Ranks

This is the part that trips up every marketing team still running the old playbook. The mechanics of AI visibility are fundamentally different from traditional search.

An Ahrefs study found that only 12% of AI citations overlap with Google's traditional top 10 search results. Put another way: 80% of the sources AI cites when answering a question do not appear in Google's top results for the same query.

What earns you a featured snippet on Google and what earns you a citation from ChatGPT are different inputs, different signals, different outcomes. Zoom's CMO Kimberly Storin told the Wall Street Journal exactly this: her SEO team, however expert, does not automatically understand what answer engine optimization requires. The skillsets are interconnected, she said, but the challenge is much bigger.

So what does AI actually cite?

Third-party sources dominate. AirOps research found that 85% of brand mentions in AI search results come from third-party pages, not from the brand's own domain. What AI tools know about your brand, and what they tell people about your brand, is almost entirely determined by what other people have written about you on other websites.

LinkedIn has emerged as the most cited domain for professional queries across every major AI platform. Research from Profound found that between November 2025 and February 2026, LinkedIn surged from outside the top 20 to the number one spot across ChatGPT, Gemini, Google AI Overviews, Microsoft Copilot, and Perplexity. A complementary study by Semrush, analyzing 325,000 unique prompts, ranked LinkedIn as the second most cited domain overall, appearing in roughly 11% of AI responses on average.

Reddit ranks as the single most cited domain overall across ChatGPT, Google AI Mode, Gemini, Perplexity, and AI Overviews combined, according to Peec AI's analysis of 30 million citations.

Review sites carry outsized weight in B2B. G2's research found that when buyers were asked what would most increase their confidence in an AI chatbot's recommendation, the top answer was a citation from a software review site.

Press releases are back. Zoom reversed course after years of deprioritizing them because AI retrieval systems parse them as structured, attributable content.

And individual experts matter more than company pages. 59% of AI citations on ChatGPT Search and Google AI Mode come from individual creators, not company pages. LinkedIn articles and posts together account for approximately 35% of all LinkedIn citations within ChatGPT responses. The most cited authors post more than five times in a four-week window. AI responses show high semantic similarity to the original LinkedIn content, meaning the exact language your people use on the platform directly shapes how AI describes your company.

Stack all of this up and the picture is clear. AI does not cite your company page. AI cites the ecosystem of content that surrounds your brand: the articles written about you, the executive posts on LinkedIn, the Reddit threads mentioning you, the press releases, the reviews, the earned media placements. The only way to influence that ecosystem at scale is to publish.

Part Four: Why Publishing Is the Strategic Response

We are not talking about content marketing. We are not talking about blog posts that rephrase your product messaging in 800 words and gate them behind a form.

We are talking about publishing. Original, expert-driven, structured content that AI systems treat as authoritative because it actually is authoritative.

The distinction matters. Content marketing asks "what information does our buyer need?" Publishing asks "what does our company believe that our competitors don't?" The first produces how-to guides any competitor could publish next week. The second produces perspectives, arguments, and original data that only your company could publish, because they come from your specific experience, your specific data, and your specific point of view.

AI rewards the second. Here is why.

AI systems are looking for citation-worthy sources. They need content they can attribute to a named expert with verifiable credentials. They need structured information they can parse and reproduce accurately. They need freshness, because content older than three months sees citation rates drop significantly. And they need depth, because shallow content does not give the model enough signal to confidently reference it.

Publishing creates the third-party citation layer. When you publish original research, data-backed arguments, and genuine expert perspectives, other people write about your findings. Journalists reference your data. Analysts cite your frameworks. LinkedIn commenters share your conclusions. That third-party coverage is what AI actually cites when a buyer asks a question about your category. You do not get third-party coverage by publishing product pages. You get it by publishing ideas worth covering.

Publishing feeds every discovery channel simultaneously. A substantive piece of original analysis does not serve one channel. It ranks in traditional search. It gets cited by AI platforms. It gets shared on LinkedIn, where AI also picks it up. It earns backlinks from other publications. It feeds email nurture campaigns. It compounds. FirstPageSage data shows that B2B SaaS posts the lowest generative engine optimization CAC of any sector at $249, precisely because the companies shipping well-structured content are the ones AI platforms disproportionately cite. Two acquisition channels, one content investment.

Publishing is the only way to control your narrative in an AI-mediated world. If you are not publishing, AI will describe your brand based on whatever it can find. That might be a three-year-old Glassdoor review. It might be a competitor's comparison page. It might be an outdated Reddit thread from someone who churned. The brands that publish control the supply of information AI draws from. The brands that do not publish are at the mercy of whatever the model scrapes together.

Part Five: The Publisher Playbook for AI Search

We have spent months studying the companies that are getting this right. The patterns are consistent.

They operate a newsroom, not a content calendar. Rippling built an editorial operation that outranked every publication covering its own story. Ramp's Econ Lab newsletter gets cited by the New York Times, Wall Street Journal, and Bloomberg because it runs on proprietary spending data nobody else has. These are not blog posts. They are editorial operations with journalist talent, editorial standards, and a publication cadence that builds authority over time.

They publish through people, not pages. The data is unambiguous on this point. Individual creators drive the majority of AI citations. LinkedIn posts from personal profiles receive 8x the engagement of identical content from company pages. 95% of B2B decision-makers say thought leadership directly influences their purchasing decisions. The companies winning AI visibility are publishing through their executives, their subject-matter experts, and their frontline practitioners, not through a faceless company blog.

They distribute on LinkedIn first. 96% of B2B content marketers already use LinkedIn to distribute content. The difference is that the companies getting cited by AI are treating LinkedIn as a primary publishing platform, not a distribution afterthought. LinkedIn articles in the 500 to 2,000 word range earn the majority of AI citations. The most cited authors post consistently. The companies getting this right are building what amounts to a LinkedIn-first newsroom that publishes at editorial speed and quality through the profiles of real people.

They maintain narrative consistency across every surface. Zoom's CMO made this point explicitly: if your messaging says one thing on your website, another on G2, another in press releases, the AI will reflect that incoherence back to buyers. LLMs synthesize across sources. Brand architecture and brand visibility are now the same thing. The companies getting this right have aligned their positioning across their website, their review profiles, their LinkedIn executive content, their press releases, and their earned media outreach. Every surface tells the same story because every surface feeds the same AI.

They measure AI visibility as a first-class metric. Only 22% of marketers are tracking AI visibility at all right now. The companies ahead of the curve are using tools like HubSpot's AEO product, Profound, and Peec AI to track brand visibility scores, share of voice, sentiment, and citation frequency across ChatGPT, Gemini, Perplexity, and Google AI Overviews. They are running prompt audits, typing the buyer queries that matter into every major AI platform, documenting where they appear and where they do not, and treating the gaps as content briefs.

Part Six: The Cost of Doing Nothing

Let us be direct about what happens if you wait.

Your competitors publish. AI learns their narrative. When a buyer asks ChatGPT "what is the best solution for X," your competitor gets named and you do not. The buyer never visits your website. They never see your ads. They never read your case studies. They form an impression of the market based on what AI told them, and you were not part of that conversation.

This is already happening. G2 found that nearly every marketer they interviewed called AI discoverability a pipeline "must-have," then admitted they were still figuring it out. For many, the wake-up call was typing a buyer query into ChatGPT and watching a competitor's name appear instead of their own.

The 95-5 rule in B2B says that only 5% of your market is actively buying at any given time. The other 95% are forming impressions. Those impressions used to be formed by trade publications, industry events, and word of mouth. Now they are being formed by AI. If you are not in the AI answer, you are not in the impression. And when that buyer enters the 5% and is ready to purchase, you are not on the shortlist.

Zoom has 99% brand awareness and still built a dedicated SWAT team for this. They are not doing it because they are paranoid. They are doing it because they read the same data we just showed you.

Part Seven: The Manifesto

We believe the following to be true.

The discovery layer has permanently shifted. A majority of B2B buyers now start their research with AI chatbots. That number will grow, not shrink. Traditional search is not dying, but it is no longer the primary front door for a growing share of the market. Any brand strategy that treats Google as the only discovery channel is a strategy built for 2020.

AI search rewards publishers. The brands that show up in AI answers are the ones that produce citation-worthy content at scale: original research, expert perspectives, structured data, genuine thought leadership. AI does not cite product pages. It cites the ecosystem of authoritative content that surrounds a brand. Building that ecosystem requires publishing, not marketing.

Every company needs to operate like a media company. Not metaphorically. Actually. With editorial talent, publication cadence, distribution infrastructure, and a commitment to saying something worth citing. The companies that build this capability will own the narrative in their category. The companies that do not will be described by whatever the model finds, if it finds anything at all.

Executive voice is infrastructure. LinkedIn is now the top cited domain for professional AI queries. The citations go to individuals, not company pages. Getting your executives to publish consistently on LinkedIn is not a nice-to-have for personal branding. It is a core distribution strategy for how AI will describe your company to every buyer who asks.

The window is open and it will not stay open. Only 22% of marketers are tracking AI visibility. Most companies have not started. The brands that move now will build citation authority that compounds over time, because AI systems tend to reinforce the sources they have already cited. First-mover advantage in AI search is real, and it is available right now.

This is a culture shift, not a budget line. Zoom's CMO said it best: "It's more of a culture shift than it is an expertise shift at this point." The companies that win will be the ones that build curiosity, publishing discipline, and cross-functional collaboration into their operating rhythm. No single team owns AI visibility. It touches content, web, data, PR, executive comms, and product marketing simultaneously.

Brand is the last satisficing signal in an AI-mediated world. When AI narrows a buyer's consideration set to two or three names, the winner is the one the buyer trusts. Trust comes from brand. Brand comes from consistently publishing things worth reading, worth citing, and worth believing. The companies investing in brand now are the ones AI will recommend later.

What Comes Next

We are going to keep covering this. Every week, we publish the data, the case studies, and the strategic frameworks that help brand leaders navigate what is happening. If you are reading this and realizing your company has not started, the good news is that most of your competitors have not either.

The bad news is that the ones who have are building an advantage that gets harder to close with every passing quarter.

Start publishing. Start now. The AI is already talking about your category. The only question is whether it is talking about you.