Most of what circulates as advice for getting cited by AI traces back to one 2023 paper, published at KDD in 2024, and a rewrite checklist that later benchmarks could not reproduce. When ten conversational-SEO methods were retested against a shared benchmark, most proved largely ineffective and often harmful to citation rank. What is left is duller than the checklist: pages an engine can fetch, evidence it can quote, and attribution it can resolve. This piece separates the two.
The short answer to what gets cited: a passage carrying a specific, sourced claim, on a page the engine can retrieve, from a source the engine already trusts. The original generative engine optimization paper improved on baseline by roughly 30 to 40% using statistics, quotations and cited sources, up to 41% for the best single method, and the published version is careful that its metric is position-adjusted visibility inside a generated answer. Phrasing tricks stacked on top of that mostly do nothing.
These are commercial stakes. G2 found AI chatbots and review sites are now the two biggest influences on B2B shortlists, and 51% of software buyers now start research with an AI chatbot more often than with Google. Getting the causal story wrong here does not cost you a ranking. It costs you a quarter of work on rewrites that testing says are inert.
What does the strongest citation evidence actually say?
It says evidence-carrying text wins, and it says so under conditions worth naming out loud. The study ran nine optimization methods over GEO-bench, a query set built for the purpose, and reported the largest gains for adding statistics, quotations and cited sources. The KDD proceedings entry and the project's own citation block both make the scope explicit.
What it measured matters more than the headline percentage. The visibility metric is position-adjusted word count inside a generated answer. Analytics links and click counts play no part in it. A gain reported as "+41% visibility" means more of your text surfaced inside a controlled answer, which is a real signal and a different one from being named or shortlisted.
The same results carry a null result that gets dropped in the retellings. Keyword stuffing, the tactic that defined a decade of optimization, produced no measurable gain: in the paper's table with error bars it scored 19.8 against a 19.8 baseline. Any checklist quoting the positive findings while omitting that one is quoting selectively. The version that says it hurts is one of the widely quoted AI search numbers that fail the check.
Why do most rewrite tactics fail when someone retests them?
Because the original gains were measured on one purpose-built benchmark, and they did not carry over to a wider test, evidence additions included. C-SEO Bench took ten conversational-SEO methods, eight from earlier work and two of its own, and tested them across two tasks and six domains, which is a wider surface than the original work covered. Almost nothing held: most methods were largely ineffective and often harmful to citation rank, and that is after correcting for running many tests at once.
The benchmark's second result is the one that should reset planning. It also measured what happens when every document in the pool applies the same method, which is the realistic case in any category where competitors read the same blog posts. Advantages that looked real in isolation shrink when everyone adopts them, and the dataset paper points instead to traditional retrieval-side improvements as the durable path.
Read practically, this is permission to stop doing a category of work. Rewriting a page in a more conversational register, or restating the same claims in question form, are not interventions that testing supports. If the rewrite does not add a fact, a source, or a retrievable answer, treat its expected effect as zero until you measure otherwise.
Which content properties actually correlate with citation?
Three are worth the work: pages the engine can fetch, evidence it can quote, and attribution it can resolve. Fix them in that order, because evidence on a page an engine cannot fetch is evidence nobody reads.
1. Pages the engine can fetch
Several major AI crawlers do not execute JavaScript. Vercel ran with MERJ a study of crawler behavior showing that the agents it measured read raw HTML, so a client-rendered page arrives close to empty. Google is the documented exception: the same study found Google's Gemini leverages Googlebot's infrastructure, enabling full JavaScript rendering, and Google's AI features only show a page as a supporting link if it is "indexed and eligible to be shown in Google Search with a snippet". Cloudflare surveyed the 200,000 most visited domains for agent readiness, which is the scale at which this question is now being asked.
2. Evidence a model can quote
Retrieval systems assemble answers out of passages, so the practical question is which passages get pulled. AirOps' March 2026 analysis of 548,534 pages ChatGPT retrieved found only 15% were cited, with pages whose titles overlapped the query by at least 50% cited 20.1% of the time against 9.3% for low overlap, and a 2026 study running 1,280 experimental conditions tested how far content-side properties carry. Read together with C-SEO Bench, the pattern is that any effect from added evidence is small and unproven on citation rank.
3. Attribution the engine can resolve
More than 4 in 10 AI citations point at sources brands do not own, so a large share of what the answer is built from sits off your domain entirely. Your own name has to resolve cleanly too: use one name and one description everywhere, so engines attribute your claims to you and not to someone else.
Does structure change what gets extracted, or just what gets read?
It changes what gets extracted, within limits, and the limits are the interesting part. Liu et al. documented that models attend unevenly across a long input context, with material in the middle used least. That finding describes the assembled context. It does not prove that mid-page evidence is penalized, though it does argue against resting a whole page on one buried passage.
Chunk size sets what a retrieval system hands the model. In a 2026 retrieval benchmark built on 1,186 multi-hop chemistry questions, the gold evidence passages averaged 188 tokens with a 56.26-token standard deviation. That describes how one dataset's evidence passages were built, not how engines chunk pages, so it is not a target length. What carries over is simpler: a claim should sit in the same passage as its own attribution. Commercial answer engines have not been measured this way.
Structure is the floor, not the differentiator. Answer-first paragraphs, question-shaped headings and clean HTML make substance extractable, and they produce nothing on a page with no substance to extract.
What should you stop doing?
Stop stuffing keywords. It produced no measurable gain in the study everyone cites for the opposite conclusion, and it degrades the same pages for human readers.
Stop reaching for persuasion framing. Scarcity and exclusivity framing measurably reduces how often a model recommends a product in controlled tests across three sets of 10 fictitious products, and the study found scarcity and exclusivity cut product visibility on every model and product it tested. Marketing register is not neutral input to a model.
Stop serving one thing to crawlers and another to people. Text hidden from users but written for machines breaks Google's spam policies, and it puts your organic presence at risk to chase a citation effect nobody has demonstrated.
Stop compressing all of this into one number. The IAB's framework splits AI visibility into Presence, Prominence, Portrayal, and Persuasion because those move independently. A single composite score hides which one you are losing, which is the only thing you needed to know.
How do you test a tactic on your own pages without fooling yourself?
Start by accepting that a single before-and-after comparison proves nothing here. The SparkToro and Gumshoe research found AI engines are highly inconsistent when recommending brands, and Search Engine Land's write-up of the same study reports that AI recommendation lists rarely repeat exactly across runs. A 2026 controlled experiment on LLM skincare recommendations shows how little it takes to flip one: with products otherwise identical, a well-known brand's edge disappeared when a competitor had a rating advantage of less than 0.1 stars.
So measure a rate, not an event. Ask each buyer question in a few phrasings over several weeks and record how often you appear, are cited, and are recommended. A 2026 variance-components study found run-to-run noise large enough to swallow real differences in small samples, and Ronald Sielinski's "Quantifying Uncertainty in AI Visibility" makes the same argument with confidence intervals. The platform medians in that work show how far apart engines sit on the same question.
Then change one thing at a time and keep a holdout. Pick ten comparable pages, add extractable evidence to five, leave five alone, and rerun the same questions on the same cadence for both sets. Without the holdout you cannot separate your edit from a model update or from ordinary drift.
Watch which sources the answers cite, alongside whether you appear at all. Seer found 87% of SearchGPT citations across a 500-citation sample matched Bing's top results at the time of the study. The source list tells you whether the fix belongs on your site or somewhere else.
How does Trovance test what actually gets you cited?
Trovance runs the protocol above as a standing system, on a repeating cycle. You define the market questions your buyers ask, and it sweeps them across AI engines every week, preserving every answer run as a snapshot with its full context: who was mentioned, who was cited, who was recommended, and which sources carried the answer. The unit of measurement is the appearance rate across many runs.
Preserved snapshots let you compare. When an answer changes, you can see whether the shift came from your page, from a third-party source that started carrying a competitor, or from ordinary run-to-run variance. Share of answer across a question set shows where you are absent entirely, which is a different problem from being present and losing the recommendation.
The Brain keeps the claims you are entitled to make with the proof behind each one, and recommended actions arrive as a ranked queue of content worth making, backed by answer evidence. Drafts are written in your voice from your author voice guidance and research sources, a person reviews and approves everything before it publishes, and the next analysis cycle reruns the same questions so you can see whether the answer moved. That is the loop the benchmark evidence supports: change something specific, then verify against current answers.
What Trovance will not do is promise a citation, a ranking, or a recommendation. No honest system can. The engines are probabilistic and the sources they trust sit outside your control, and a retest in which most tactics failed or backfired is what the testing record looks like for tactics sold as certainties. There is no universal visibility score here either, and we are building toward broader engine and source coverage. That work is in progress.
What should you do this week?
Work the three properties in order. Fetch your five most important pages with JavaScript disabled and read what comes back, because that is what the crawler read. Then audit those pages for extractable evidence: specific numbers and named studies, each dated and attributed to a primary source that fifty other pages have not already quoted.
Then check the sources the answers actually cite for your buyer questions, and see how many of them you have any presence in at all. If the answers are built on review sites and community threads that never mention you, no amount of on-page rewriting reaches that surface, and the work belongs in earned coverage instead.
Finally, set the measurement up before you change anything, so this quarter's edits produce evidence. If you want the runs, snapshots and comparisons handled for you, start a free Trovance trial and see which buyer questions you are absent from, and which sources are carrying the answers instead.
Read the evidence correctly
The most-quoted study in AI search, quoted correctly - what the GEO paper measured and what it did not.
The GEO playbook for getting cited by AI engines - what is left once you remove the tactics that failed retesting.
Why more evidence can still produce worse AI answers - adding facts is not linear, and the failure mode is specific.
AI agents need evidence, not more content - why publishing volume stopped correlating with being used.
Test it on your own pages
How to track AI citations - building an appearance rate you can actually act on.
How to measure AI search visibility without one score - the dimensions a composite number hides.
Do headings help AI retrieve long documents? - the structural question, answered with retrieval research.
A relevant page can still miss the proof - relevance and evidence are separate failures with separate fixes.
