In a recent LinkedIn post, Ross Hudgens, CEO of Siege Media, argued that Perplexity’s market share is shrinking. So, including it alongside ChatGPT, Gemini, Claude, and Google’s AI search products can give marketers a distorted picture of where their visibility actually matters. Ross declared, “Everyone should remove Perplexity from their LLM trackers today.”
I think Hudgens is asking the right question. I am less convinced that he has the right answer.
The AI search market is clearly concentrating around a handful of major platforms. But it has not reached the point where SEO professionals can confidently focus on the “four major models” and stop looking elsewhere.
In fact, the data suggests we may be heading toward a three-tier market rather than a settled leaderboard.
I’ve Seen This Movie Before
This argument triggered a flashback for me.
In March 2002, I attended Search Engine Strategies in Boston. The dot-com crash was still fresh, and I was working for an SEO agency in Greater Boston that tracked our clients’ rankings across 15 different search engines.
Someone at the conference argued that we should simplify our reporting and concentrate on the five major search engines of the day, Yahoo, Excite, Lycos, AltaVista, and Ask Jeeves.
I pointed out that Google was missing from the list.
That turned out to be an important omission. Google was gaining search share rapidly, even though it was not yet one of the obvious incumbents that everyone agreed belonged on the list.
That experience makes me cautious about declaring any emerging search market “settled.”
The question isn’t simply which platforms are the biggest today. It is which platforms are big enough, growing fast enough, or strategically important enough that ignoring them could cause an SEO to miss the next shift.
See also: AI Search Isn’t Replacing Google, It’s Layering On Top – Similarweb Data
Perplexity Really Is Losing Ground
There is plenty of evidence supporting Hudgens’ argument.
StatCounter’s worldwide AI chatbot referral data shows a dramatic change during 2026. In June, Perplexity accounted for 7.91% of AI chatbot referral share, essentially tied with Gemini at 7.94%. By August, Perplexity had fallen to 4.31%, while Gemini had risen to 10.9%.
The distinction between referral share and usage matters here.
Similarweb’s latest publicly available comparison tells a different story about the size of the individual platforms. Its May 2026 data put ChatGPT at 53.9% of worldwide web visits among seven major AI assistants, Gemini at 27.9%, Claude at 9.2%, DeepSeek at 4.1%, Grok at 2.4%, and both Perplexity, and Copilot at 1.3%.
That makes Hudgens’ concern about weighting understandable.
Suppose an SEO platform gives Perplexity the same weight as ChatGPT even though Perplexity represents only a tiny fraction of the traffic in the relevant market. A particularly strong or weak result on Perplexity could indeed distort the aggregate score.
But there is a problem with going from that observation to “remove Perplexity.” Small market share does not necessarily mean small strategic importance.
See also: Google Gemini Sends More Traffic To Sites Than Perplexity: Report
The Market Is Consolidating, But Not Around One Winner
The biggest change in 2026 is not the disappearance of Perplexity. It is the emergence of a serious second player.
Similarweb reported that ChatGPT’s share of AI chatbot web traffic fell from 76.4% a year earlier to about 52.7% in May 2026. Gemini rose from roughly 9% to 27.3%, while Claude increased from 1.6% to 8.9%.
That is a significant redistribution of attention. But it does not mean ChatGPT is disappearing.
OpenAI reported it had more than 900 million weekly active users in February, and said at the end of July that its models reach more than one billion active users across all its products.
Google, meanwhile, reported in August that the Gemini app had passed 1 billion monthly users.
And Claude has established a particularly strong position in enterprise and developer workflows. Anthropic said in April that more than 100,000 customers were running Claude on Amazon Bedrock. In August, Anthropic told investors its annualized revenue had passed $65 billion and that enterprise adoption continued to grow.
Those aren’t three versions of the same business. They represent three different distribution advantages.
ChatGPT has enormous direct consumer adoption.
Gemini has Google’s distribution across Search, Android, and its broader product ecosystem.
Claude has developed a substantial position in enterprise and professional work.
That looks less like a winner-take-all market than an emerging oligopoly.
And Then There Is Google
This is where I think LLM tracking gets conceptually messy.
Should Google AI Overviews and AI Mode be treated as another LLM? For SEO purposes, I would argue no. They are more important than that.
Google reported in June that AI Overviews were reaching more than 2.5 billion users per month, while AI Mode had surpassed 1 billion monthly users.
Google also said AI Mode queries had more than doubled every quarter since launch.
Similarweb data reported by TechCrunch showed AI Overviews appearing in 43% of Google U.S. searches by May 2026, up from 15% a year earlier. AI Mode visits increased from 126 million in June 2025 to 279 million in May 2026.
Those numbers make one thing clear. If your objective is understanding how people discover information, products, companies, and websites through AI, Google’s AI search experience cannot be treated as a niche LLM. It is an AI layer inside the world’s dominant search ecosystem.
That makes AI Overviews and AI Mode a separate tracking category.
Claude Is The Reason Usage Share Alone Is Not Enough
Claude presents another problem with Hudgens’ approach.
If an SEO team ranks platforms strictly according to consumer traffic, Claude looks relatively small. But that could be misleading for B2B companies.
Anthropic has been aggressively building enterprise distribution. PwC said it would deploy Claude Code and Cowork across its global workforce and train and certify 30,000 professionals. TCS announced a plan to make Claude available to 50,000 employees across 56 countries. Cognizant has expanded its Claude partnership and says more than 30,000 associates have completed Claude training.
Anthropic also says more than 1,000 business customers were spending more than $1 million annually as of April 2026.
That creates an important distinction for SEO and content marketers. The platform that sends the most referral traffic today isn’t necessarily the platform that deserves the most attention from your company. A B2C retailer, a SaaS company, and a cybersecurity vendor may need three different LLM tracking portfolios.
So, has the market consolidated?
My answer is yes, but only partially. There is enough evidence to stop pretending that every AI assistant deserves equal weight.
The data increasingly supports a core group consisting of ChatGPT, Gemini, and Claude, with Google AI Overviews and AI Mode treated as a separate, and arguably larger, search-discovery layer. Perplexity belongs in a different category.
It is no longer reasonable to treat Perplexity as a peer to ChatGPT or Gemini in an overall market-share calculation. But I would not delete it from the tracker. I would de-weight it.
That distinction matters.
Perplexity’s own trajectory is worth watching. Similarweb’s August 2026 analysis still described Perplexity as an active player in the AI search market, including its strategy of remaining ad-free while pursuing subscriptions and enterprise deals. Similarweb also partnered with Perplexity to make its market intelligence available inside Perplexity Computer.
That doesn’t prove Perplexity will regain share. It does demonstrate why “small” and “irrelevant” aren’t synonyms.
What SEO Professionals Should Track Now
I would organize AI visibility measurement into three tiers.
Tier one: Should be the platforms that have both scale and strategic significance. Track ChatGPT and Gemini separately, rather than averaging them into a generic LLM score. Add Claude for B2B, enterprise, and professional audiences.
Tier two: Should be AI search embedded in established ecosystems. Track Google AI Overviews and AI Mode separately from ChatGPT, Gemini, and Claude. The goal here isn’t merely measuring an LLM. It is measuring how AI is changing the search journey. Microsoft Copilot may also belong in this tier for organizations with substantial Microsoft 365 exposure. Microsoft says its first-party Copilot family has more than 150 million monthly active users, while its broader AI features across Microsoft products reach 900 million monthly active users.
Tier three: Should be emerging or specialized platforms. This is where Perplexity, Grok, DeepSeek, and other platforms belong.
Don’t give them equal weighting. Don’t ignore them either. Instead, monitor them for unusual visibility, referral traffic, citations, and growth.
Don’t Let The Average Hide The Signal
There is one part of Hudgens’ argument that I strongly agree with. An aggregate LLM visibility score can create a false sense of precision.
Imagine that a brand has a 40% citation rate on ChatGPT, 35% on Gemini, 30% on Claude, and 90% on Perplexity. A simple average produces 48.75%.
But that number doesn’t tell a marketer much.
If Perplexity accounts for only a small percentage of the audience or referrals that matter to the business, its 90% score should not have the same influence as the ChatGPT result.
The better question is not “What is our average LLM visibility?” It is “Where are the people who matter to this business actually encountering our brand?”
That requires connecting three datasets.
First, measure audience exposure. Look at usage, visits, and platform distribution.
Second, measure visibility. Track mentions, citations, linked URLs, and the prompts that produce them.
Third, measure business impact. Connect AI referrals to engagement, leads, sales, and other conversions wherever the analytics allow it.
That produces a much more useful measurement model than a simple leaderboard.
My Verdict On Perplexity
So, I think Ross is right about the problem. I don’t think he’s quite right about the prescription.
Don’t remove Perplexity from your LLM tracker. Remove it from the position of equal weighting.
That’s an important distinction.
If Perplexity represents 1% of your AI referrals, giving it 25% or 33% of your visibility score is indefensible. If it represents more than 5% of your AI referrals, as Hudgens himself suggests, the case for keeping it becomes much stronger.
And if a particular vertical gets disproportionate visibility from Perplexity, deleting it could erase a useful competitive signal.
The same principle applies to every new AI platform that comes along.
That is the lesson I took from that conference in Boston in early 2002. The mistake wasn’t tracking too many search engines. The mistake was assuming that today’s market leaders would still be tomorrow’s market leaders.
Google was easy to overlook because it wasn’t yet one of the five engines everyone had agreed mattered.
Two decades later, the lesson is still relevant.
Track the leaders. Weight them according to reality. But keep watching the outsiders that are gaining ground.
Because in search, the platform you stop tracking may be the one you wish you hadn’t.
More Resources:
- Perplexity AI Interview Explains How AI Search Works
- Data Shows AI Citation Patterns Reveal Strategic SEO Opportunities
- LLM Guidance Doesn’t Transfer The Way SEO Guidance Did
Featured Image: Gorodenkoff/Shutterstock
