For instance, Digiday reports that Buzzstream, using the Xofu tool, analyzed 4 million AI citations across ChatGPT, Google's Gemini, AI Overviews, and AI Mode in January. Their findings showed news publications accounted for 14% of all citations, primarily in evaluative or decision-making queries. Conversely, Similarweb identified a different mix of top publishers, focusing on how often a brand name or source was mentioned, not just linked.
This lack of standardization means "nobody knows anything yet," according to Vince Nero, director of content marketing at Buzzstream. He draws a parallel to the early days of SEO, emphasizing that expertise in this new domain is still nascent. The immediate implication is that publishers cannot rely on a single metric to gauge their influence in the AI era.
One crucial distinction highlighted by Digiday is between "citations" (a direct link to a publisher's site) and "mentions" (referencing a brand name or source in an AI summary). Reuters, a wire service, was the only publisher to appear in analyses from Buzzstream, Similarweb, and Profound, suggesting a unique advantage for rapid, factual reporting in AI search. However, even for top publishers, AI visibility has been volatile over the past nine months, a fluctuation attributed by Adelle Kehoe, Similarweb's director of product marketing, to the news cycle and ongoing adjustments to large language models (LLMs).
The most striking revelation in the evolving AI landscape is the disconnect between high visibility in AI responses and actual referral traffic. While some publishers are frequently cited, the number of clicks they receive from LLMs can be surprisingly low. For example, Similarweb data indicates The New York Times receives the most AI referral traffic, despite ranking eighth in Similarweb’s AI visibility index. Conversely, Reuters, which ranks first in AI visibility, is ninth in terms of referral traffic. This demonstrates that being seen doesn't always translate into reader engagement on a publisher's site.
Publishers are, however, adapting. Alphonse Hardel, head of Reuters’ news agency, attributes their frequent AI mentions to the "speed, accuracy, and breadth" of their reporting. Reuters has also secured content licensing deals with major LLMs, including Microsoft and Meta, acknowledging that while AI citations are growing at a "very low base," they align with their mission of disseminating factual information. Hardel also highlighted that AI search helps them build business relationships with AI companies.
The Washington Post has taken a proactive approach, signing deals with OpenAI, Google, Amazon, and Perplexity, and hiring a head of AI discovery. According to Karl Wells, The Washington Post’s chief revenue officer, people referred from LLMs spend more time on their site and convert to subscriptions at a rate four to five times higher than traditional search referrals. This highly engaged audience is crucial for their advertising and subscription businesses. Meanwhile, Business Insider has reported year-over-year growth in visits and return visits, even as Google's AI Overviews have reduced traffic for many other publishers, underscoring the potential for strategic engagement.
The broader market trend suggests a shift away from traditional search as a primary traffic driver. Condé Nast CEO Roger Lynch anticipates a future where Google search is "no longer a meaningful driver" for their business, despite increased revenue. This perspective, along with the fact that LLMs currently drive only about 1% of overall web traffic, suggests that AI visibility is more about brand authority and direct content licensing opportunities than a new channel for mass referral traffic.
For Developers
Focus on building tools that help content creators track AI mentions and citations across various LLMs, not just links. The complexity of current measurement suggests a market need for unified, robust analytics platforms. For Founders: Consider direct content licensing deals with AI companies as a viable revenue stream and brand building strategy, rather than solely relying on organic AI referral traffic, which remains low. For Publishers: Prioritize deep user engagement and high-quality content that encourages direct visits and subscriptions. Data suggests AI-referred users are more engaged, making content excellence a powerful lever for conversion, as seen with The Washington Post's subscription rates. For Tech Professionals: Understand that AI's impact on content discovery is still nascent. The current "messiness" offers opportunities for innovation in content distribution, audience analytics, and AI-publisher partnerships. Frequently Asked Questions Why is AI visibility data inconsistent for publishers? The inconsistency stems from varied methodologies used by analytics firms, including different AI platforms scanned, types of user prompts, and how "mentions" versus "citations" are defined, leading to divergent results. Does being cited in AI chatbots guarantee more traffic to a publisher's site? No, data shows a clear misalignment. While publishers may appear frequently in AI responses, this high visibility does not consistently translate into increased referral traffic to their websites, with LLMs contributing only about 1% of overall web traffic. How are publishers adapting to the challenges of AI content surfacing? Publishers are adapting by signing direct content licensing deals with AI companies, hiring specialists for AI discovery, and focusing on high-quality, engaging content that fosters deeper user interaction and subscription conversions, as demonstrated by The Washington Post. Research Sources digiday.com
Even though a publisher is frequently cited in AI responses, this doesn't guarantee more clicks to their website because being seen in AI summaries doesn't always translate into reader engagement. For example, The New York Times receives the most AI referral traffic despite ranking lower in AI visibility, while Reuters ranks first in AI visibility but lower in referral traffic.
Publishers are adapting by re-evaluating their strategies beyond traditional SEO, and some are exploring direct licensing deals with AI companies. The Washington Post, for example, sees higher engagement from AI referrals, demonstrating that adapting strategies can yield positive results.
AI visibility data for publishers is inconsistent across analytics firms due to differing methodologies. These discrepancies arise from the specific AI platforms analyzed, the types of prompts used, and how "mentions" or "citations" are defined, making it difficult for publishers to pinpoint their actual reach.
In the context of AI, a 'citation' refers to a direct link to a publisher's website, while a 'mention' refers to referencing a brand name or source within an AI summary. Reuters, for example, uniquely appears in analyses from multiple firms due to its rapid, factual reporting, giving it an advantage in AI search.
More insights on trending topics and technology
