How SEMRUSH Went Invisible to ChatGPT and Fought Back: Inside Battle Story for AI Visibility
December 29, 2025
Just weeks after Semrush launched Enterprise AIO and the AI Visibility Toolkit, an internal test delivered an uncomfortable surprise. A simple query was entered into ChatGPT asking about AI monitoring tools. The response listed every major competitor in the space. Semrush was missing.
For a company built on search intelligence, the omission was jarring. Despite a recent product launch and strong brand recognition in SEO, large language models had no awareness of Semrush as a player in AI monitoring and visibility. That discovery exposed a deeper and more troubling reality about how influence now works in AI-driven search.
At the same time, Semrush noticed another contradiction. ChatGPT and other LLMs were citing Semrush blog content hundreds of times, yet organic traffic to those articles was declining. The brand was present as a source but absent as a recommendation. Citations created reach without authority, visibility without positioning, and content value without measurable impact.
That disconnect forced a fundamental rethink of SEO strategy in the age of generative AI. Using its own technology, Semrush built a structured framework to measure and improve LLM visibility. Within one month, its AI share of voice for key prompts jumped from 13 percent to 32 percent. The shift marked a turning point in how the company now understands search, content, and brand influence.
The Measurement Crisis Behind AI Search
Traditional SEO metrics rely on clicks, rankings, and conversions. AI search breaks that model. A user can see a brand recommended inside ChatGPT or Perplexity and never visit a website. Influence happens without traffic.
Semrush could confirm that LLMs were using its content, but that data alone did not reveal whether the brand was being positioned as a solution. Without insight into competitive placement, investment decisions became guesswork.
The operational challenge was even greater. Classic SEO assumes rankings are relatively stable and can be checked weekly. AI platforms are non-deterministic. Studies show that between 40 and 60 percent of sources cited by LLMs change every month. The same prompt can generate different answers within hours.
It became clear that new metrics were required, metrics designed to track influence rather than visits.
Read Also:
The Two Metrics That Redefined AI Visibility
After testing numerous approaches, Semrush narrowed AI influence down to two indicators that consistently reflected real impact.
Visibility measures whether a brand is mentioned at all for a specific prompt. It is a simple yes or no. Either the brand exists in the AI conversation or it does not.
Share of voice measures how a brand is positioned when mentioned. Appearing first in a list of recommendations carries far more weight than appearing last.
Together, these metrics function like traditional SEO signals, but they measure authority and presence rather than clicks. Semrush tracked them across ChatGPT, Perplexity, Google’s AI features, and other platforms using Enterprise AIO.
Prompt selection proved critical. Broad queries such as “AI tools” offered little insight. The focus shifted to bottom-funnel prompts that signal buying intent, including comparisons and best-in-class questions. When a brand appears in those answers, it enters the buyer’s consideration set at the most decisive moment, without paying for ads.
The Framework Semrush Used to Fight Back
The first step was identifying a tightly controlled set of high-intent prompts that reflected real decision-making behavior. These were not keyword-stuffed phrases but natural questions prospects would actually ask AI systems.
Next came baseline measurement. Early tracking confirmed the problem. Semrush held just 13 percent share of voice in AI visibility conversations, validating the concern that LLMs did not associate the brand with its new offerings.
Tracking moved from weekly to daily. Because AI responses fluctuate constantly, performance had to be viewed in ranges rather than fixed positions. Stability, a core assumption of classic SEO, no longer applied.
Content audits followed. Semrush reviewed existing blog posts, guides, and product pages to identify natural opportunities to connect AI visibility topics directly to Enterprise AIO. Rather than forced promotion, the goal was clarity. Content already discussing the problem now clearly presented the solution.
The most significant strategic change came when Semrush expanded beyond its own website. LLMs draw from a wide ecosystem that includes Reddit, Quora, forums, social platforms, and trusted third-party domains. Semrush coordinated efforts across community discussions, acquired media assets, and social channels to ensure consistent brand presence wherever AI systems gather information.
New content was then created with citation in mind. Clear definitions, structured explanations, and data-backed statements performed best. Direct answers replaced creative flourish. Precision replaced metaphor. The goal was simple: make content easy for both humans and machines to understand and reuse accurately.
What the Results Revealed
Within a single month, Semrush’s AI share of voice for AI visibility topics rose from 13 percent to 32 percent. Visibility for non-brand industry queries also increased, proving the framework extended beyond branded searches.
The speed was unexpected. Improvements appeared within days, sometimes within hours, far faster than traditional SEO timelines.
The volatility introduced a new urgency. With citations changing monthly, outdated content can no longer wait in a backlog. Drops in visibility now demand immediate response.
Revenue attribution remains complex. Separating the impact of AI visibility from paid search, email, and other channels is difficult, but the influence is undeniable.
What stood out most was the effect of off-domain optimization. Expanding beyond semrush.com amplified brand mentions far more than on-site changes alone.
What This Means for the Future of SEO
AI-driven search is already reshaping discovery. Top-funnel traffic will continue to decline as users receive direct answers from LLMs. Measuring visibility and share of voice is now as important as tracking clicks once was.
Websites alone are no longer sufficient. Brands must be present wherever AI systems look for expertise, including forums, communities, and social platforms.
Leadership teams must be prepared for success that does not immediately appear in analytics dashboards. Influence often happens before traffic, not after.
Content teams must adapt to faster cycles. In AI search, delayed updates mean lost visibility.
The lesson from Semrush’s experience is clear. The shift to AI-driven discovery is happening now, not in the future. Teams that begin measuring and optimizing for LLM visibility today will shape the next phase of search. Those who wait for certainty will be competing in a landscape that has already moved on.