The question that haunts every SEO professional in 2026 is deceptively simple: where does your brand show up when someone asks an AI instead of typing into Google? The answer, as Steven Coufal discovered while running SEMrush’s AI citation tracking for both a fintech company and a local med spa, is often surprising and occasionally embarrassing. In a conversation with Thibaut de Lataillade, Coufal walked through what the citation data actually reveals, why a competitor’s dormant LinkedIn page was outperforming dedicated websites in AI results, and why the economics of AI referral traffic might be more favorable than the click-through numbers suggest.
What Semrush’s Citation Tracking Actually Shows
SEMrush’s AI search tracking tool monitors citations across three engines: ChatGPT, Gemini, and Google’s AI overviews. It groups queries topically, so for Coufal’s work at Retired.com, it surfaces a “retirement advice” cluster with every query that falls within that category. For each query, the tool shows which sites are being cited, which are not, and what the competitive landscape looks like inside the AI-generated results.
The interface is straightforward enough. You see your citations, you see your competitors’ citations, you see the gaps. But the value is not in the dashboard itself – it is in the specific, actionable discoveries that emerge when you actually look at what the AI engines are pulling in and from where.
Coufal is candid about the tool’s limitations. Data availability is the biggest weakness – the coverage is not comprehensive, and the engines themselves are evolving faster than any tracking tool can keep up with. But for getting a baseline understanding of where you stand and where the obvious gaps are, he considers it more than sufficient to justify the effort.
The LinkedIn Discovery Nobody Expected
Between his role at Gartner Digital Markets and his current position at Retired.com, Coufal did some freelance work for a local med spa in Austin. The business ranked well for core terms in traditional search but wanted to understand its AI search visibility. What SEMrush revealed was not what either party anticipated.
A competitor’s LinkedIn company page was being pulled into AI search citations. Not a website. Not a blog. A LinkedIn page with irregular posting and no particularly recent content. But it was filling a content gap – a specific informational need that no other source in the competitive landscape was addressing – and the AI engines were citing it because it was the best available answer for that particular query.
“I initially told them, I was like, yeah, I don’t think you guys really need to bother with LinkedIn. I don’t think that’s really a big channel for you.”
Coufal had to reverse his own advice. The data showed clearly that LinkedIn, a platform he had dismissed as irrelevant for a local service business, was functioning as a citation source for AI search engines. The lesson is not that every business needs a LinkedIn strategy. The lesson is that AI engines are pulling from sources that SEO professionals have historically ignored, and the only way to know which sources matter for your specific competitive landscape is to look at the data.
The Outdated Profile Problem
The second discovery was more prosaic but equally actionable. SEMrush surfaced professional profiles for the med spa’s doctor – profiles on medical directories, review sites, professional networks – that had been filled out a decade or more ago and never updated. These profiles contained outdated addresses, old contact information, and stale biographical details.
The problem was not that these profiles existed. The problem was that AI search engines were citing them. When someone asked ChatGPT or Gemini about this doctor or his practice, the AI was pulling information from these outdated sources and presenting it as current. The brand narrative being constructed by AI engines was being shaped by profiles the doctor had forgotten existed.
The fix was trivially simple – an hour of work updating the profiles with current information. But the diagnosis required the citation tracking tool to surface what was happening. Without it, nobody would have known that a Healthgrades profile from 2012 was actively shaping how AI search engines described the practice in 2026.
The broader implication for any business is worth stating plainly: AI engines do not just cite your website. They cite every source they can find that addresses the query. If you have outdated profiles, abandoned social media pages, or stale directory listings, those sources are candidates for citation – and you will have no idea until you actively monitor what the AI engines are doing with them.
The Economics of AI Referral Traffic
Coufal raises a point that reframes the entire AI search conversation from doom-and-gloom to cautiously optimistic. Research from G2 suggests that while AI search results generate fewer click-throughs than traditional organic results, the clicks that do come through convert at significantly higher rates.
The logic is intuitive once you hear it. A user who asks an AI engine a detailed question, reads the synthesized answer, and then clicks through to a cited source has already qualified themselves. They are not browsing. They are not comparison-shopping at the top of a funnel. They have a specific need, the AI answer addressed it partially, and they want more depth from the original source. That is a fundamentally different – and more valuable – click than someone scanning ten blue links and picking the one with the most appealing title tag.
For commercial keywords in particular – the queries that actually drive revenue – Coufal argues this shift is manageable. B2B purchases, financial decisions, any transaction that involves meaningful money: people are not going to complete those inside an LLM. They are going to click through to the source, evaluate it, and convert there. The total click volume may decline, but the per-click value may increase enough to compensate.
The practical challenge remains measurement. Coufal admits he does not have a reliable way to track click-through rates from AI citations specifically, and he openly asks if anyone else does. Google Analytics can show referral traffic from AI sources, but attributing specific conversions to specific AI citations is still more art than science. What you can do – and what Coufal recommends – is monitor which commercial keywords you are appearing for in AI citations and track whether referral traffic from those sources is converting. The signal is there. The attribution is imperfect. But waiting for perfect measurement before acting on visible trends is a luxury that growth marketers have never been able to afford.
For the full interview breakdown, see our complete Expert Insight with Steven Coufal.
Tools Mentioned in the Interview
The following tools and platforms were referenced during this conversation.
Related Content
From Link Building to AI Workflows: Expert Insights with Steven Coufal
Steven Coufal has spent over 15 years scaling organic growth from compliance content to international SEO empires to AI-powered workflows. He breaks down how Claude and SEMrush are replacing day-long research sprints, why chasing Google updates is a losing game, and what LLM citation tracking actually reveals about your brand.
Stop Chasing Google Updates: Why a User-First SEO Strategy Wins
Steven Coufal explains why obsessing over Google algorithm updates is a recipe for professional burnout, how an SEO consultant helped him shift to a user-first strategy, and why building for users instead of algorithms is the only approach that survives the fracturing search landscape.
How Claude and SEMrush Power a Modern SEO Content Workflow
Steven Coufal walks through the exact workflow he uses to combine Claude, SEMrush, and Ahrefs into a content production system that compresses day-long research sprints into hours while producing output that is genuinely competitive and unique.