Originally published on The Searchless Journal
Every new search channel goes through the same arc. It opens wide, rewards early movers with disproportionate visibility, and then gradually fences off the easy wins until what remains requires budget, expertise, or both. AI search is in the middle of this arc right now, and the window for earning citations without spending is measurably closing. The brands that recognize the pattern have a narrow advantage. The brands that do not will find themselves on the outside of a fence they did not see being built.
The comparison most worth studying is not the last algorithm update. It is the history of press release SEO, a channel that went from wide-open to fenced-off overnight, and the agencies that adapted are the ones that survived.
The Press Release SEO Precedent
From roughly 2003 to 2013, distributing optimized press releases through wire services was one of the most effective SEO tactics available. It worked because it stacked three payoffs simultaneously. First, keyword-rich anchor text in the release generated direct ranking signals for the client's site. Second, the wire service's own domain authority meant the release itself often ranked for the target term, driving referral traffic. Third, journalists reading the wire could pick up the story and write original coverage with organic links back to the client.
Three benefits from one action, all for the cost of a wire distribution fee. It was the kind of arbitrage that defines an early channel.
On July 30, 2013, Google updated its Link Schemes guidelines and explicitly classified optimized anchor text in press releases as unnatural links. The first payoff disappeared overnight. Every client relying on that anchor text for direct SEO lift lost it in a single update. What did not disappear was the referral traffic from distribution, and the indirect benefit when reporters wrote original stories. The agencies that had built their strategy around all three payoffs adapted. The ones that depended entirely on the anchor text arbitrage did not.
AI search is running a version of the same play, and the fencing has already started.
The Current State Of The Opening
A Fuel Online audit of 1,000 enterprise domains found that 62 percent are technically invisible to AI models. Ask ChatGPT, Perplexity, or Gemini an unbranded question about their category, the kind a buyer actually types, and the models fail to mention them 81 percent of the time. This is the size of the opportunity. Most of the field has not shown up yet. The brands that do show up, even minimally, are winning citations in a low-competition environment.
The economics of this window are stark. Profound's data puts the median time to first citation for new content at 6.81 days. Get published, get retrieved, get cited, all within a week. The barrier to entry is speed and content quality, not budget. There is no paid placement in AI citations yet, no equivalent of Google Ads that lets you buy your way into the answer. The citation economy is still earned, which is what makes it both valuable and temporary.
Shane Tepper, co-founder of Resonate Labs, describes the window in competitive rather than temporal terms. The timeline is not a date on a calendar. It is however long it takes your competitors to wake up. In crowded B2B categories, that clock is already running. Brands that established citation presence in the first half of 2026 are building the topical authority and entity recognition that AI engines rely on for source selection. Every month that passes, more competitors enter the space, and the marginal cost of earning a citation rises.
The May 7 Inflection Point
The structural shift in AI referral traffic that began on May 7 of this year has held. Similarweb reported a 157.7 percent jump in ChatGPT referral traffic on that date, and subsequent analysis confirmed it was a new baseline rather than a spike. Profound tracked the same structural doubling across every brand basket it monitors. Three separate measurement firms confirmed the shift using independent methodologies.
What this means is that the channel is not just growing, it has stepped up to a new plateau. More users are clicking through from AI answers to source sites. The referral traffic that brands earn from citations is now structurally higher than it was in the first quarter. This makes each citation more valuable, which makes the closing window more urgent.
The parallel to press release SEO is direct. In the early days, wire distributions generated massive referral traffic because the releases ranked directly in search results. As Google's algorithm evolved and the wire services' own rankings declined, the direct traffic payoff weakened. The same dynamic will play out in AI search. As more brands optimize for citations, the share of voice per brand declines. As AI engines refine their source selection, the criteria for inclusion will tighten. The window narrows from both ends.
What Closing Looks Like
The fence in AI search will not arrive as a single algorithm update the way Google's 2013 link schemes action did. It will arrive through several mechanisms, some already in motion.
First, source selection criteria will tighten. AI engines currently cite a relatively wide range of sources because the citation economy is immature. As these systems collect more data on source quality, accuracy, and user engagement, they will narrow the set of sources they trust. Brands that are not already in the trusted set will find it harder to break in.
Second, paid placements will arrive. ChatGPT already runs advertising. Perplexity runs ads. When these platforms introduce sponsored citations or paid inclusion in AI answers, the earned citation will lose share of voice to paid placements. This is the same dynamic that played out in traditional search, where paid results pushed organic results below the fold.
Third, content saturation will raise the barrier. The brands publishing GEO-optimized content today are competing against a relatively thin field. As more brands adopt GEO practices, the volume of citation-worthy content in any given category will multiply. The AI engines will have more sources to choose from, and citation shares will fragment.
How To Use The Window While It Is Open
The brands gaining ground right now share several characteristics. They publish frequently, with original data and expert analysis rather than rewritten commodity content. They structure their content for AI extractability, using clear headers, factual claims with verifiable sources, and entity-rich language. They maintain presence across multiple AI engines rather than optimizing for a single platform. And they track their citation performance with the understanding that single measurements are noisy, using repeated queries to establish reliable baselines.
The most important characteristic is speed. The window rewards action over analysis. A brand that publishes ten well-structured articles this month and starts accumulating citations has a meaningful head start over a competitor that spends three months building a strategy document. The strategy matters, but the citations are what compound. Each cited article builds the entity recognition and topical authority that makes the next citation easier to earn. The flywheel is already spinning for early movers. It has not yet started for the 62 percent of enterprise domains that remain invisible.
The lesson from press release SEO is not that early arbitrage disappears. It is that the agencies and brands that used the arbitrage period to build durable assets, real audience, real media relationships, real brand authority, survived the fencing-off and kept winning. The ones that relied entirely on the arbitrage did not. The AI citation window is the same opportunity. Use it to build something that lasts after the fence goes up.
Top comments (0)