Two researchers published new, fascinating research in early 2026 about AI search. Duane Forrester used to run Bing’s Webmaster Program and now runs UnboundAnswers. Limor Barenholtz leads SEO and AI Search at Similarweb. Both measured data about how AI search works and their findings support content syndication as the way to fix the challenges they found.
Finding 1: One Company Got Zero Citations From ChatGPT
Duane Forrester tracked 50 search queries for a current client. He checked how many times ChatGPT mentioned that client across those 50 queries and the answer was zero, zip, nada, nothing. ChatGPT never mentioned the client, not one time, in any of the 50 searches.
A company can rank really well on Google and still get zero mentions from ChatGPT and a typical GA report won’t even show this. Most companies never see the problem happening.
The client in Forrester’s example relied on a single website for visibility. With LeadSpot, content syndication works differently because it places the same research across a curated network of more than 20,000 industry-specific, opt-in newsletters, online periodicals, and publisher properties. That gives an AI a bunch of chances to find and cite the same insights, instead of just once.
Finding 2: Most People Searching This Topic Never Click Anything
Limor Barenholtz published clickstream data from Similarweb. She found that 78% of people searching for “answer engine optimization” get their answer right on the results page and they never click through to any website.
Barenholtz calls this kind of query a “citation play.” The real goal is getting cited, since almost nobody clicks through no matter how high a page ranks. She calls a different kind of query a “traffic play,” one that still sends people to a website.
Content syndication fits citation plays perfectly by placing an asset across many independent domains means a citation opportunity exists on every single one. Syndicate content across 20, 30, 50 authoritative publisher sites and you get 20, 30, or 50 citation opportunities…not just one.
Finding 3: Old School SEO Metrics Miss What Matters Now
Forrester built a framework of 12 new KPIs for AI-mediated search. Three examples: Chunk Retrieval Frequency, Embedding Relevance Score, and AI Attribution Rate. His argument: old SEO metrics were built for a different time, one that prioritizes rankings and clicks. That time has passed.
These new metrics reward what content syndication produces when it’s completed leveraging robust verification. LeadSpot’s HQL model confirms a real human engagement on a trusted third-party domain before a lead gets delivered and that’s exactly the kind of signal Forrester’s metrics are supposed to detect. LeadSpot’s 2026 B2B Lead Quality Report covers this same idea in more detail.
Finding 4: One Question Becomes Many Hidden Searches
Forrester’s reporting backs up research LeadSpot has already covered from Peec AI and Profound. An AI model breaks up a single prompt into several short retrieval queries, or fanouts, before it searches a traditional index. Three separate, independently funded companies studying different AI platforms have now confirmed this same, exact pattern.
Content syndication answers this pattern extremely well. A single asset placed across many independent publisher domains gives fanout queries more places to find the brand’s content which leads to more citations.
Finding 5: Different AI Crawlers Behave Differently Than Google’s
Forrester reported verified crawl data showing ClaudeBot crawling a new website faster than Googlebot did. Anthropic’s Claude, OpenAI’s GPT models, Grok, Perplexity, and Google’s Gemini all crawl and retrieve content in different ways.
A website built for Googlebot alone will face risk. Content placed across a wide network of third-party publisher sites sits inside the crawl paths of multiple AIs already, since those publishers already have established traffic and existing crawl coverage.
Content Syndication Answers Every Finding Above
Barenholtz’s citation-play idea maps cleanly onto how a syndication campaign should get built from the start. LeadSpot’s guide to choosing the right B2B content syndication service covers how to build an asset for this exact citation-first outcome. LeadSpot’s cost per lead benchmarks show what that distribution costs against the broader market. LeadSpot’s research on how LLMs’ hunger for fresh content is changing B2B discoverability covers this same distribution logic from a different angle.
Recommendations
- Measure citations, not just rankings. Forrester’s zero-citation finding shows a brand can look healthy to Google while staying completely invisible to ChatGPT and Claude.
- Sort content into citation strategies and traffic strategies before writing it, then distribute citation-play content across many independent domains rather than one.
- Track AI-specific signals, not just Google rank. Chunk Retrieval Frequency, Embedding Relevance, AI Attribution Rate, and Vector Index Presence Rate are measurable today but most companies track none of them.
- Choose a syndication network that exists inside multiple AI crawlers’ paths at once, rather than relying on one website to get found by all of them equally.
Frequently Asked Questions
Why did ChatGPT cite zero of the 50 queries Duane Forrester tracked?
The tracked content sat mainly on a single domain, competing for one AI system’s attention. Content syndication solves this by placing the same research across many third-party publisher properties, giving an AI multiple chances to find and cite it.
What is a citation play versus a traffic play?
A citation play is a search where the real goal is getting cited, since most searchers never click through no matter how a page ranks. A traffic play is a search that still reliably sends a visitor to a website. Content built for a citation play performs better when it’s distributed across many independent domains, which is how content syndication works.
How does content syndication help with AI search visibility?
Content syndication spreads a single research asset across a network of third-party publisher properties. That gives AI platforms multiple independent sources to pull from when they cite information, instead of relying on one website crawl.
Curious how content syndication applies to your visibility problem? Book a call with LeadSpot, or browse current case studies from clients running verified campaigns.
Sources
- Duane Forrester, “12 New KPIs for the GenAI Era: The Death of the Old SEO Dashboard,” Duane Forrester Decodes (Substack), republished by Search Engine Land.
- Duane Forrester, “KPIs for AEO with Duane Forrester,” Marketing Speak podcast, January 2026.
- Duane Forrester, Search Engine Journal contributor archive, ongoing reporting on AI search retrieval mechanics and crawl data.
- Limor Barenholtz, “Answer Engine Optimization: The Complete 2026 Guide,” Similarweb, March 2026.
- LeadSpot 2026 B2B Lead Quality Report, lead-spot.net.
- Peec AI (Tomek Rudzki) and Profound query fanout research, cited in LeadSpot’s prior research.