Close-up of a person's hand holding an iPhone and using Google AI Mode, an experimental mode utilizing artificial intelligence and large language models to cess Google queries, Lafayette, California, March 24, 2025.
(Smith Collection | Gado | Photos | Getty ImagesMany believe generative artificial intelligence engines and large language models are a blem for the open web, publishers and even advertisers.
Some think they are a blem for Google because they'll have less traffic to monetize, replaced by Gemini LLM conversations.Both of these notions are wrong.In just a single year since launch, Gen AI and LLMs have changed the face of the open web.
This rapid change is impacting both engines and publishers. We're seeing a decline in traditional traffic, with some publishers experiencing more significant drops than others.
This trend directly affects Google, which is seeing fewer queries and also impacts publishers who are receiving less referral traffic from engines.
The competition for this shrinking pool of traffic is heating up, with companies OpenAI and Perplexity vying for a — and Google actively competing with itself by offering a generative AI duct called Gemini.Whether OpenAI and ChatGPT "kill" Google or Google itself does it, there's no doubt– traffic to publishers is going down.
The question is what do publishers do if traffic drops 20%, 30% or even 50%?
It is ly the end of as we know it and, counterintuitively, the best thing that could have ever happened to Google and the open web.Gen AI is reshaping and rebuilding the publisher modelGen AI tools ChatGPT, Perplexity and Google's Gemini are stealing traffic from publishers.
Data suggests that just 0.1% of traffic from these engines are making their way to publishers, meaning if you're ing on ChatGPT or Perplexity the lihood you then click out to read more a topic is next-to-nil.
In fact, the newest buzzword from publishing executives is 'zero click es' whereby not a single action is taken after a is performed.No publisher is immune to this threat.
referral traffic to 1,000 web domains decreased by nearly 7% year over year, nearly a billion critical visits that have disappeared from the open web.
On its face, you would think there is no rational defense from Gen AI companies that are tanking publisher traffic.
Google would try to put a spin on things by publicly claiming that people who click on links after seeing AI Overviews tend to spend more time on those sites.
David Tramontan | SOPA Images | Lightrocket | Getty ImagesBut the Gen AI threat is causing a chain reaction for the foundation of publishers.
Today, the primary determining factor to how much revenue a publisher makes is tied to how many potential people view ads on their site. Put simply, less traffic means less revenue in the short term.
This has made it necessary for publishers to rethink how they make money. It's even made Google rethink how it makes money.
Short-term disruption is inevitable, but what's coming next is a leap toward deeper engagement and bigger revenue than we've ever seen before — both for Google and to the open web.
The rise of high-value LLM conversationsWhile traffic will decline, potentially hurting Google and publishers in the short term, a new type of "supply" is emerging: LLM conversations.Nearly a billion people use ChatGPT, Gemini or Perplexity each month.
Most are asking general knowledge questions, but an estimated 10%–20% of those chats are high-value: travel planning, health care, commerce, financial advice and more.
These conversations are worth far more than a click.ChatGPT's model is straightforward, earning $20 per month from each r, making every question valued the same, whether you asked the weather or a trip to Cancun.
Google's Gemini, backed by a $200 billion ads and 10 million advertisers, will see these conversations as commercial opportunities. Over time, a multiday exchange a trip can evolve.
If a person asks a question, Gemini can reply in the moment or if more re is needed, it can come back with more ideas, until that conversation ends with a full-service transaction.
From here, users can book flights, hotels, and restaurants.This is an incredible revolution for Google, from each interaction worth $1 per click to $1,000 per conversation.
This shift positions Gemini's long-term revenue potential well beyond . In essence, less traffic — but a lot more revenue. Publishers have an equally powerful opportunity.
While 30%–40% of their audience today arrives via , 60%–70% comes direct.
By launching their own Gemini-style LLMs, publishers can host high-value conversations within their trusted environments, covering travel, shopping, health and finance.
This is where deep content and expert voices can guide meaningful decisions.While ChatGPT can instantly give you ideas and information a trip you're considering, you would never book it on the spot.
You'll still seek out reviews from people who've been there, browse images and s and explore different perspectives before making the decision.
While AI offers a new way to , the human need remains the same–to be informed and educated on what matters before making a decision in a way that creates trust and confidence.
Publishers have expertise. They have earned trust over decades. They can use that to create amazing value in this new landscape.This is where publishers win over the next decade.
Major purchases and commitments require more than quick answers; they require exploration of reviews, imagery, s, and authoritative reporting, all of which publishers already der.Some publishers will monetize these interactions through transactions I believe Gemini will, others through subscriptions ChatGPT is doing now.
For those who lean in, this represents a once-in-a-generation shift from pageview-driven models to relationship-driven value.
User trust becomes the foundation for both resilience and growth in the LLM era. The future of publishers (and Google) is bright.Adam Singolda is founder and CEO of Taboola.