The debate surrounding AI search visibility measurement reached a critical inflection point following a provocative public declaration by Siege Media CEO Ross Hudgens. In a widely discussed LinkedIn post published in September 2026, Hudgens asserted that Perplexity’s shrinking market share invalidates its inclusion alongside heavyweights like ChatGPT, Gemini, Claude, and Google’s native AI search products. His core thesis is straightforward: grouping Perplexity with dominant platforms distorts aggregate visibility metrics, leading marketers to misallocate resources based on skewed performance dashboards. Hudgens went so far as to urge digital marketers and search engine optimization (SEO) professionals to immediately purge Perplexity from their large language model (LLM) trackers.
While Hudgens correctly identifies a persistent vulnerability in how modern SEO dashboards calculate aggregate scores—namely, the risk of distortion through unweighted averaging—his prescriptive remedy oversimplifies a rapidly maturing and structurally fragmented search ecosystem. Deleting emerging or niche search channels from tracking software risks blinding organizations to strategic shifts, much like dismissing nascent platforms in the early 2000s would have cost digital marketers vital foresight. Rather than discarding platforms outright, the industry requires a sophisticated, tiered measurement architecture that weighs traffic, strategic utility, and platform distribution accurately.
The Evolution of Search Metrics and Historical Precedents
To understand the current anxiety over AI search fragmentation, it is helpful to examine historical precedents in digital marketing. The debate over which platforms merit inclusion in tracking software mirrors discussions from the post-dot-com era. In March 2002, search engine optimization practitioners gathered in Boston for the Search Engine Strategies conference. At the time, agencies routinely tracked client visibility across up to 15 distinct search engines. Amid pressure to streamline reporting dashboards, industry voices argued for consolidating metrics around the five dominant incumbents of the era: Yahoo, Excite, Lycos, AltaVista, and Ask Jeeves.
Notably absent from that proposed consensus list was Google, which was rapidly gaining query share despite not yet possessing the entrenched market dominance of the older portals. Omitting Google from tracking portfolios proved to be a costly strategic error for early search marketers who failed to anticipate structural market shifts. This historical parallel underscores a fundamental rule of digital discovery: the primary objective of tracking is not merely cataloging today’s largest monopolies, but identifying platforms whose growth trajectories or unique audience distributions make ignoring them a strategic risk.
Examining the Data: Perplexity and the Shifting Traffic Landscape
An objective analysis of current web analytics and market research substantiates Hudgens’ premise that Perplexity has surrendered significant ground to larger competitors over the course of 2026. According to worldwide AI chatbot referral data published by StatCounter, Perplexity experienced a sharp contraction during the summer months. In June 2026, Perplexity commanded approximately 7.91 percent of AI chatbot referral share, putting it in a virtual dead heat with Google Gemini, which held 7.94 percent. By August 2026, however, Perplexity’s referral share plummeted to 4.31 percent, while Gemini surged ahead to capture 10.9 percent of total referral traffic.
Similarweb’s comprehensive market research on global web traffic and visits to AI assistant platforms provides additional context regarding raw platform scale. Publicly available metrics from May 2026 position OpenAI’s ChatGPT as the dominant market leader, capturing 53.9 percent of worldwide web visits among the seven major AI assistants analyzed. Google Gemini followed in second place with 27.9 percent, Anthropic’s Claude secured 9.2 percent, DeepSeek claimed 4.1 percent, Grok registered 2.4 percent, while both Perplexity and Microsoft Copilot lagged at 1.3 percent each.
These stark disparities explain why visibility weighting concerns are valid. When SEO analytics software aggregates visibility scores across disparate platforms without adjusting for actual audience volume, a brand holding a near-perfect citation score on a low-traffic platform like Perplexity can artificially inflate its overall AI visibility rating. Consequently, a company might believe its comprehensive LLM visibility stands at an impressive 50 percent, while the vast majority of real-world users interacting with AI systems are doing so via ChatGPT or Google.
The Consolidation of the AI Search Market
Rather than moving toward a single, monopolistic winner, the AI search and assistant market has organized itself into an oligopoly characterized by distinct distribution moats. Historical data from Similarweb highlights a dramatic redistribution of user attention over a twelve-month period leading up to mid-2026. In the prior year, ChatGPT maintained an overwhelming 76.4 percent share of AI chatbot web traffic. By May 2026, that share receded to approximately 52.7 percent as users diversified their toolsets. Over the same timeframe, Gemini expanded its footprint from roughly 9 percent to 27.3 percent, and Claude climbed from 1.6 percent to 8.9 percent.
Despite losing some of its monopoly-like market share, OpenAI continues to scale aggressively. In February 2026, OpenAI reported exceeding 900 million weekly active users, a figure that crossed the milestone of one billion active users across its product ecosystem by late July. Google reported parallel milestones for its consumer ecosystem, announcing in August 2026 that the Gemini application had surpassed one billion monthly active users, with internal data showing that 63 percent of users actively utilized voice interaction capabilities.
Meanwhile, Anthropic has carved out a distinct and highly lucrative market position by targeting enterprise and developer workflows. Announcements from Anthropic throughout 2025 and 2026 highlighted major deployments with global professional services and technology firms, including PwC, Tata Consultancy Services (TCS), and Cognizant, which integrated Claude code and enterprise solutions across tens of thousands of professionals globally. By July 2026, Anthropic reported to investors that its annualized revenue run rate had surpassed $65 billion, driven heavily by enterprise adoption where traditional consumer traffic metrics fail to capture the platform’s true economic impact.
The Integration of AI Within Established Search Ecosystems
Any comprehensive evaluation of AI search visibility must also account for native search integrations, particularly those deployed by Google. Treating Google AI Overviews and AI Mode as standard, standalone LLMs obscures their massive scale and foundational impact on the modern digital journey. Google reported in June 2026 that AI Overviews were reaching more than 2.5 billion users per month, while AI Mode queries had surpassed one billion monthly users, doubling every quarter since its commercial launch.
Third-party validation from Similarweb and analysis published by TechCrunch indicate that AI Overviews appeared in 43 percent of U.S. Google searches by May 2026, representing a nearly threefold increase from 15 percent the previous year. Concurrently, visits to AI Mode climbed from 126 million in June 2025 to 279 million in May 2026. These figures demonstrate that Google’s AI search features are not niche chat interfaces; they represent a fundamental evolution of the world’s most dominant search distribution network. Similarly, Microsoft Copilot continues to wield substantial influence, leveraging integration across the Microsoft 365 enterprise suite to reach over 150 million monthly active users for first-party Copilot applications and up to 900 million users across broader ecosystem AI features.
A Tiered Architecture for Modern AI Visibility Measurement
Given the complexities of consumer adoption, enterprise distribution, and ecosystem integration, a binary decision to include or banish specific models is inadequate. Industry analysts and SEO strategists advocate for a structured, three-tier framework for tracking AI visibility:
Tier One: Core Consumer and Enterprise Leaders
This tier encompasses platforms with immense scale and distinct strategic footprints. ChatGPT and Gemini must be tracked independently rather than collapsed into an homogenized LLM score. Claude should be included in this primary tier specifically for organizations targeting B2B, developer, and enterprise professional audiences, where traditional consumer web traffic metrics understate its commercial relevance.
Tier Two: Ecosystem-Embedded AI Search
This category isolates search experiences natively integrated into established digital pathways, primarily Google AI Overviews, AI Mode, and Microsoft Copilot. The objective here is not merely measuring conversational LLM engagement, but quantifying how generative artificial intelligence alters the traditional user search journey and referral pathways.
Tier Three: Emerging, Specialized, and Alternative Platforms
This tier accommodates platforms such as Perplexity, Grok, and DeepSeek. Rather than assigning them equal weight—which creates statistical distortion—or deleting them entirely, marketers should monitor these channels for specialized vertical visibility, citation patterns, and nascent growth trends. Perplexity, for example, continues to secure strategic partnerships, such as its collaboration with Similarweb to integrate market intelligence directly into Perplexity Computer, proving that a smaller consumer footprint does not equate to zero strategic value.
Implications for Enterprise Analytics and Strategy
Aggregating disparate AI visibility data into a single, unweighted average score introduces dangerous statistical noise into marketing dashboards. If a brand maintains a 40 percent citation rate on ChatGPT, 35 percent on Gemini, 30 percent on Claude, and 90 percent on Perplexity, a simple unweighted average yields 48.75 percent—a figure that fails to reflect reality if Perplexity accounts for only one percent of actual brand discovery channels.
A robust measurement model requires connecting three distinct datasets: audience exposure metrics to understand platform distribution, visibility analytics to track brand mentions and citations across prompts, and business impact metrics to tie AI-driven referrals to actual user engagement and conversions.
Ultimately, the debate over Perplexity’s place in LLM trackers highlights the maturation of AI optimization as a professional discipline. De-weighting minor channels prevents metric distortion, but maintaining visibility monitoring across all active platforms ensures that digital strategists remain prepared for the next structural shift in how users find information online.
