GEO Claims a Conversion Lead Over SEO as AI Rewires Brand Discovery

Generative engine optimization (GEO), the practice of structuring content so AI platforms like ChatGPT, Google Gemini, and Perplexity cite it in their answers, is producing measurably different results from traditional search engine optimization (SEO) in first real-world conversion tests. A South Korean duty-free retailer running an Enliple GEO pilot recorded purchase conversions 63% above its conventional search baseline, with AI-driven traffic growing at 37.3% versus 21.9% for keyword-driven traffic. Gartner projects traditional search volume will fall 25% this year. Nearly 60% of Google searches already end without a click to any external website.

Two decades of SEO infrastructure, built around backlinks, keyword density, and metadata calibrated for Google’s PageRank algorithm, don’t translate to how large language models (LLMs, the AI systems behind tools like ChatGPT and Gemini) decide what to cite. That structural gap is what marketing teams in Seoul, Tokyo, and London are now spending money to close.

The Search Traffic That Stopped Coming Back

What Zero-Click Searches Are Costing Publishers

Google search referrals to publishers fell 33% globally in the year to November 2025, per Similarweb data. Individual sites absorbed far steeper losses: HubSpot estimates it lost 70 to 80% of its organic traffic; Chegg, the education platform, reported a 49% decline; DMG Media documented drops as steep as 89% for some queries.

• 60% of all Google searches end without a click to an external site (Similarweb)
• 83% zero-click rate when a query triggers a Google AI Overview; 93% in Google AI Mode (Seer Interactive, Semrush)
• 33% decline in Google search referrals to publishers, year to November 2025 (Similarweb)
• 25% projected drop in traditional search volume this year (Gartner)

Informational queries absorb the worst of it. Ahrefs data shows 99.9% of informational keywords now trigger a Google AI Overview, and Ahrefs research found the number-one ranked result loses roughly half its click-through rate when one appears. A brand can hold the top position on Google and still lose half of what that position used to deliver in traffic.

Why Backlinks Don’t Help With LLMs

For 25 years, Google’s PageRank logic rewarded inbound links from authoritative domains. Agencies built businesses around link acquisition, domain authority scores, and outreach programs designed to earn citations from high-authority sites. That moat is intact for traditional search; it does almost nothing inside an LLM’s citation logic.

Researchers at Princeton University, Georgia Tech, and the Allen Institute for Artificial Intelligence formalized this distinction in a 2023 academic paper that coined the term. LLMs evaluate brand authority through different signals: entity mentions across third-party publications, forums, and communities; structured schema markup on the brand’s own site; and content formatted for direct machine extraction. An unlinked brand mention in a Reddit thread can directly influence whether an LLM cites a brand; paid backlinks from high-domain-authority sites register as essentially irrelevant to that citation decision.

Research attributed to Princeton found that fewer than 10% of sources cited in AI-generated answers matched the top-10 Google organic results for the same queries. Holding Google’s top rankings carries almost no predictive power for AI citation.

The practical implication is a fundamental repricing of what SEO agencies sell. A link-building campaign with well-documented effects on Google rankings has near-zero documented effects on LLM citation rates. The same applies to meta description tuning, title tag optimization, and structured keyword placement; all of it is calibrated for a crawler that scores relevance through mechanisms an LLM doesn’t replicate.

ChatGPT Flips the Ad Switch

OpenAI’s advertising pilot launched in the United States on February 9, 2026. Within six weeks, per ppc.land’s documentation of the ChatGPT advertising rollout, the pilot crossed $100 million in annualized revenue. Canada, Australia, and New Zealand joined March 26. On June 6, Benji Shomair, OpenAI’s VP of Monetization, confirmed the UK had gone live, the first time OpenAI served ads outside those four markets. Japan, South Korea, Brazil, and Mexico follow in coming weeks.

Ads are restricted to ChatGPT’s Free and Go tiers only. Users on Plus, Pro, Business, Enterprise, and Education plans see no ads. Categories in the pilot include shopping, retail, and travel. David Dugan, formerly at Meta, heads global ads solutions for OpenAI. Ad-tech partners include StackAdapt, Kargo, Pacvue, Criteo, and Adobe.

OpenAI is projecting $2.5 billion in advertising revenue for 2026, per Adweek figures, while operating losses are expected to reach $14 billion for the same year according to figures Digiday reported in April. Advertising is financial necessity, not optional growth.

South Korean marketers were watching closely. When KPMG Korea published its May 28 report, “The Reshaping of the Content Industry by AI and Growth Strategies,” it named GEO as “emerging as a new core competitive advantage for companies” and stated businesses would “increasingly need to design content not only for human users but also for AI comprehension and reasoning systems.” ChatGPT ads going live in South Korea mean organic AI citations will sit next to paid placements inside the same interface, raising the cost of absence for brands not present in either.

Conversions Up 63% in a Duty-Free Pilot

Enliple, a South Korean data-tech company, ran a GEO pilot with a major domestic duty-free retailer before launching its i-GEO product commercially in May 2026. Post-implementation, conventional search traffic was up 21.9%. AI-driven traffic outpaced it at 37.3%. Purchase conversions linked to actual transactions rose 63%, and AI traffic inflow finished 9 percentage points above competing duty-free operators over the same period.

The numbers from Akamai Technologies’ own internal test run in a different direction entirely. Akamai, a US CDN (content delivery network) and cybersecurity company, piloted its AI Brand Presence product on its global website before the commercial May 19 launch. By creating a parallel AI-readable version of the site and compressing machine-readable data loads by 99%, Akamai’s AI Brand Presence investor release reported an 85% increase in citations and a 364% surge in brand presence in general searches where Akamai wasn’t mentioned by name. ChatGPT presence relative to competitors rose 133%.

The assumptions underpinning 20 years of digital strategy are breaking down in real time. We’re witnessing one of the most significant shifts in digital history, a move from human-driven search to machine-mediated discovery.

Patrick Sullivan, CTO of Security Strategy at Akamai Technologies, in the company’s AI Brand Presence press release, May 19, 2026.

The Vendors Building the GEO Stack

South Korea’s commercial GEO market compressed most of its activity into a two-week window in May. Plus Zero launched its Gewriter platform on May 13, positioned as an on-site GEO engine that restructures brand websites for AI parsing without requiring full architectural rebuilds. PK Communication, which describes itself as a dedicated GEO consulting firm, offers diagnostic roadmaps across seven categories including AI crawler permissions, structured data markup, and external citation signals. Enliple’s i-GEO scores AI citation potential, optimizes site structure, and auto-generates content for LLM extraction.

Naver, South Korea’s dominant search platform, plans to publish guidelines later this year to help bloggers and content creators in key verticals improve their likelihood of being cited by AI systems, extending the GEO push beyond brand advertisers into the broader creator economy.

The global GEO market sits at approximately $1.09 billion in 2026, according to Dimension Market Research’s GEO market sizing report, and is projected to grow at a 40.6% compound annual growth rate through 2034. Samsung Securities, in a May 28 analyst note, described “marketing strategies historically centered on keyword advertising and SEO” as “now rapidly expanding into AI response optimization and GEO.”

Why the GEO Market Can’t Measure Its Own Results

GEO has no single dominant analytics platform. Boston Institute of Analytics noted this in a June 2026 guide to the discipline: unlike SEO, which has Google Search Console, Semrush, and Ahrefs as standardized audit tools, GEO optimization runs primarily on vendor dashboards and post-hoc citation checks, with no cross-platform standard for what counts as a meaningful AI citation.

Citation logic operates across multiple separate agents. GPTBot (OpenAI), PerplexityBot, ClaudeBot (Anthropic), and Google-Extended crawl sites independently. A brand’s robots.txt controls which of those agents see its content. A site that has inadvertently blocked GPTBot can run a full GEO program and never surface in a ChatGPT answer; most brands discovered that risk only after auditing their technical access settings.

Terminology adds more friction. GEO and AEO (answer engine optimization) are used interchangeably across the market, with no standard definition separating the two. Commercial tools use different proprietary metrics: Akamai measures “brand presence” via its own dashboard; Enliple tracks AI traffic inflow against conventional search baselines; PK Communication scores against seven diagnostic categories. No cross-platform citation-authority score exists the way domain authority scores exist for traditional SEO.

The 2023 Princeton, Georgia Tech, and Allen Institute paper that coined the GEO term proposed that the “mention graph,” how often a brand is referenced across third-party sources, forums, and communities, is rapidly replacing the link graph as the signal LLMs read for authority. Unlike PageRank, which SEO practitioners spent years reverse-engineering, the weighting of that mention graph inside any given LLM is proprietary and shifts with each model update, leaving optimization programs chasing a target they can’t fully observe.

Akamai observed a 300% annual increase in AI bot traffic across enterprise sites. Seer Interactive’s 2026 data found brands cited in Google AI Overviews earn a 35% higher organic click-through rate than those not cited. The performance case for GEO is building on real deployment data. The measurement infrastructure to verify that data independently still hasn’t arrived.