Generative Engine Optimization is a real and developing field focused on how information becomes visible and cited inside generated answers. Academic research has measured meaningful effects and production evidence confirms it can operate at substantial scale. The evidence also reveals a significant limitation: techniques that improve content at one stage can actively reduce visibility elsewhere in the retrieval and generation pipeline.

Generative Engine Optimization has developed quickly from an academic research problem into a commercial category. Agencies now offer GEO services, software companies are building products that track citations and mentions inside generated answers, and companies are beginning to consider which teams should be responsible for visibility in these systems.

The research behind Generative Engine Optimization is real. The more difficult question is how much of the advice now being sold under the GEO label has actually been demonstrated to work.

The distinction matters because generative systems do not simply read a webpage and decide whether to quote it. Information may first have to be discovered, retrieved and reranked before it reaches the model responsible for constructing an answer. Research published in 2026 found that optimization techniques that appear useful when researchers examine only the final generation stage can perform differently when the complete retrieval process is included. (SAGEO Arena; 2026 critical survey)

Generative Engine Optimization therefore represents a legitimate area of research, but it is not yet a settled science with a universal set of optimization rules. (2026 critical survey)

What Is Generative Engine Optimization?

Generative Engine Optimization describes efforts to improve how source information performs inside systems that retrieve information and use generative models to construct answers.

The term was formally introduced in the paper "GEO: Generative Engine Optimization," submitted to arXiv in November 2023 by Pranjal Aggarwal, Vishvak Murahari, Tanmay Rajpurohit, Ashwin Kalyan, Karthik Narasimhan and Ameet Deshpande. The researchers introduced Generative Engine Optimization alongside a benchmark called GEO-bench and tested multiple methods for changing source content to determine whether those changes affected visibility in generated responses. (original GEO paper)

Their results established that the way information is written and presented can influence how prominently it appears after a generative system has access to it. The paper reported visibility improvements of up to approximately 40 percent under its experimental conditions. (original GEO paper)

That result became one of the most frequently repeated numbers associated with Generative Engine Optimization, but the number requires context. The approximately 41 percent relative increase on the paper's Position-Adjusted Word Count metric came primarily from Quotation Addition. Statistics Addition also performed strongly, producing an improvement of roughly 31 percent on the same metric, while the combination of Statistics Addition and Fluency Optimization produced the strongest combined result in the paper's testing. (original GEO paper)

Those findings establish that particular content changes affected visibility in the experimental environment. They do not establish that adding quotations or statistics to a webpage will produce the same percentage increase across live generative systems. A 2026 critical survey of the field specifically cautioned that the foundational results were conditional on a source already being present within a fixed context and therefore did not establish organic discoverability or durable traffic effects. (2026 critical survey)

Does Generative Engine Optimization Actually Work?

The research shows that Generative Engine Optimization techniques can affect visibility, citation and source usage, but there is not yet evidence that a single technique reliably improves performance across platforms and over time. (original GEO paper; 2026 critical survey)

The original GEO experiments produced several meaningful results beyond the headline visibility figure. One of the most notable involved the Cite Sources method, which affected documents differently depending on their existing position. Sources ranked fifth in the underlying search results experienced a 115.1 percent increase in visibility under that method, while sources already ranked first experienced a 30.3 percent decline in the same experiment. (original GEO paper)

The finding suggests that GEO methods may not simply increase visibility uniformly. Their effects can depend on where a source begins and how the generative system evaluates competing information.

Keyword stuffing produced a different result. It did not provide a meaningful overall improvement and was associated with an approximately 10 percent decline specifically in the Perplexity portion of the original evaluation. (original GEO paper)

The scope of that result is important. The research does not establish a universal 10 percent penalty for keyword stuffing across every generative engine. It shows that a familiar search-manipulation technique failed to provide a meaningful overall GEO benefit and performed worse in one measured portion of the experiment. (original GEO paper)

Taken together, the early research demonstrated something important but narrower than many commercial descriptions of Generative Engine Optimization imply. Content can be optimized in ways that affect generated answers, but those effects vary by technique, source position, domain and system. (original GEO paper; 2026 critical survey)

What Did SAGEO Arena Reveal About Generative Engine Optimization?

SAGEO Arena found that some Generative Engine Optimization techniques perform substantially worse when researchers test the entire path from retrieval to final citation instead of assuming that the optimized document has already been retrieved. (SAGEO Arena)

Published in February 2026, SAGEO Arena addressed a structural limitation in much of the earlier GEO research. Previous benchmarks commonly began with predetermined candidate documents. That allowed researchers to study what happened during generation, but it abstracted away the retrieval and reranking stages that determine whether a document reaches the generator in the first place. (SAGEO Arena)

SAGEO Arena created a more realistic testing environment using 171,003 documents and 2,700 queries. The system incorporated real web documents and structural information, including schema markup, while evaluating what happened as content moved through retrieval, reranking and generation. (SAGEO Arena)

The results introduced an important caution into the Generative Engine Optimization literature. Body-text-only optimization reduced average presence in the top 20 retrieval results by approximately 9 percent. Presence in the top 10 after reranking fell approximately 16 percent, while final citation declined approximately 6 percent. (SAGEO Arena)

The paper's own description of this dynamic is direct: body-text optimization "not only fails to improve generation-stage visibility but actively degrades retrieval performance, causing optimized documents to drop out of the retrieval results and never reach the generator." (SAGEO Arena)

The significance of those numbers is not simply that one optimization technique performed poorly. They show why the stage at which Generative Engine Optimization is measured matters. A document can be rewritten in a way that appears more favorable once it reaches a generative model while simultaneously becoming less competitive during retrieval. If the document falls out of the candidate results before generation begins, improvements designed for the final answer stage have no opportunity to help it. (SAGEO Arena)

SAGEO Arena also found that structural information could help mitigate these limitations and concluded that effective optimization needs to account for different stages of the pipeline rather than treating generation as an isolated event. (SAGEO Arena)

Can GEO Optimization Reduce Search and Generative Visibility?

Yes. The SAGEO Arena results provide evidence that some forms of optimization can reduce retrieval and final citation when evaluated through a more realistic search and generation pipeline. (SAGEO Arena)

This is one of the most important distinctions between experimental Generative Engine Optimization research and simplified commercial advice. A live generative search system may make several decisions before an answer is produced. It has to identify relevant information, retrieve candidate sources, determine which candidates deserve stronger placement and decide which information should ultimately be used or cited in the generated response. (SAGEO Arena; 2026 critical survey)

Optimizing only for the final stage can therefore create an incomplete strategy. A publisher might make a passage more suitable for citation while inadvertently changing the document in a way that reduces its ability to survive retrieval or reranking. Under those circumstances, the content may theoretically be better suited to the generator but practically less likely to reach it. (SAGEO Arena)

The current evidence does not establish that body-text optimization is inherently harmful. It establishes that body-text optimization alone cannot be assumed to improve the complete pipeline and that, under SAGEO Arena's testing conditions, it measurably reduced performance at multiple stages. (SAGEO Arena)

That is a substantially more cautious conclusion than the claim that Generative Engine Optimization can be accomplished through a standard list of writing techniques.

What Does the Scientific Research Say About GEO in 2026?

The broader research supports Generative Engine Optimization as a legitimate field while showing that researchers have not yet identified a stable technique that reliably improves organic visibility across different platforms and over extended periods. (2026 critical survey)

A critical survey published in July 2026 reviewed 45 GEO-related studies conducted across the field's development. The survey concluded that already-retrieved content can causally alter its citation or use, but it found no technique in the reviewed literature demonstrating a stable, longitudinal, cross-platform causal effect on organic discoverability or downstream behavior. (2026 critical survey)

That evidence boundary is central to understanding the current state of Generative Engine Optimization. It would be inaccurate to conclude that GEO does not work, because controlled research has shown that optimization can change source visibility and citation behavior. (original GEO paper; 2026 critical survey) It would also be inaccurate to conclude that researchers have established a universal GEO formula that publishers can apply with predictable results across generative platforms. (2026 critical survey)

The survey found that topical relevance and context position were among the most reproducible factors across the literature and warned that generic optimization heuristics transferred poorly between environments. (2026 critical survey)

A separate finding in the same survey analyzed 252,000 trials across six different language models and eighteen factors. That research identified relevance and position as the primary determinants of whether a source receives a citation at all, redirecting attention toward the upstream discovery and retrieval stages rather than content polish alone. (2026 critical survey)

Separate 2026 research examining Google Search, Gemini and AI Overviews provides another reason for caution about universal optimization rules. Using a benchmark of 11,500 real-user queries, researchers found substantial differences between the sources retrieved by traditional Google Search, Gemini and AI Overviews, with average source overlap below 0.2 by Jaccard similarity. The researchers also found that AI Overview results were less consistent across repeated runs and less robust to minor changes in query wording. (Google AI Overviews study)

The implication is significant. A source can be highly visible in one discovery environment and absent from another, even when those systems appear to answer the same basic question.

What Is the Difference Between SEO, AEO and GEO?

Search Engine Optimization, Answer Engine Optimization and Generative Engine Optimization address overlapping parts of the same information-discovery process, but they are not interchangeable terms.

Search Engine Optimization is primarily concerned with whether information can be discovered, understood and ranked by search systems. Answer Engine Optimization focuses more specifically on whether information can efficiently satisfy a question. Generative Engine Optimization extends the problem into systems where retrieved information must survive additional selection and generation processes before becoming part of a generated answer. (SAGEO Arena; 2026 critical survey)

A webpage may be technically accessible and competitive in traditional search yet poorly structured for a system attempting to extract a concise answer. Another page may contain an excellent answer but fail to achieve sufficient retrieval visibility to reach the generative system that could cite it.

Generative Engine Optimization therefore does not necessarily replace SEO or AEO. Current research instead supports viewing them as related parts of a broader discovery problem. SEO concerns whether information can be found and ranked. AEO concerns whether information can effectively satisfy a question. GEO concerns whether that information can survive the larger retrieval, reranking and generation process and ultimately contribute to a generated response. (SAGEO Arena; 2026 critical survey)

Treating those disciplines as completely separate can obscure the fact that generative systems still depend on upstream information discovery.

Does Schema Markup Matter for Generative Engine Optimization?

Research from SAGEO Arena indicates that structural information such as correctly implemented schema markup can matter within a generative retrieval pipeline, particularly when combined with content optimization. (SAGEO Arena)

The finding requires careful interpretation. SAGEO Arena did not establish schema markup as a universal shortcut to generative visibility. It found that structural information could help mitigate limitations associated with optimization strategies that otherwise degraded retrieval and reranking performance. (SAGEO Arena)

This suggests that machine-readable structure and written content can perform complementary functions. Text communicates information to readers and systems, while structural markup provides additional information about what a document contains and how its components relate to one another.

The research therefore provides a stronger basis for treating Generative Engine Optimization as an information-architecture problem rather than reducing it to rewriting paragraphs for language models. That distinction is particularly important because much of the commercial discussion around GEO has concentrated on content modifications. Full-pipeline research suggests that the technical structure surrounding that content deserves comparable attention. (SAGEO Arena; 2026 critical survey)

Is There Real Production Evidence That GEO Can Work at Scale?

Yes. A team of Pinterest researchers published a production case study reporting approximately 20 percent organic traffic growth from a large-scale system they describe as Generative Engine Optimization. (Pinterest GEO study)

The Pinterest example is important because it moves the discussion beyond controlled benchmarks. The published framework operated across billions of images and tens of millions of collections. Its architecture used vision-language models to predict what people would search for, AI agents to identify emerging demand, multimodal embeddings to create semantically coherent collection pages and authority-aware internal linking across large numbers of visual assets. (Pinterest GEO study)

The researchers reported approximately 20 percent organic traffic growth from the deployed system, contributing to multi-million monthly active-user growth. (Pinterest GEO study)

The case study does not prove that Pinterest's methods will produce the same result for another website or business. It does provide production evidence that a sophisticated Generative Engine Optimization system can operate at substantial scale and contribute to measurable organic growth.

The architecture is also revealing because it looks considerably different from the simplified version of GEO often presented as content optimization. Pinterest's approach involved understanding demand, organizing enormous information collections, predicting relevant searches, improving semantic relationships and building internal authority structures. (Pinterest GEO study)

In that environment, Generative Engine Optimization begins to resemble retrieval engineering and information architecture as much as conventional content optimization.

Who Is Viral Nation and Why Does It Matter to Generative Engine Optimization?

Viral Nation is an established social media and creator marketing company that has expanded into Generative Engine Optimization through a service it calls AI Discovery. Viral Nation describes the service as helping companies understand, improve and measure how AI systems describe, cite, trust and recommend their brands. (Viral Nation)

Its importance to the developing GEO category comes from the direction in which it is approaching the problem. While much of Generative Engine Optimization research and commercial practice has concentrated on webpages, search infrastructure and content retrieval, Viral Nation enters the field from an existing business built around creators, social intelligence, audience behavior and cultural signals.

Viral Nation reports that its broader social operation drove more than 85 billion creator views in 2025 and that its SocialAI infrastructure processes more than one million posts daily. Those are company-reported figures describing Viral Nation's existing social business, not independently verified measurements of the effectiveness of its AI Discovery or GEO service. (Viral Nation)

That evidence boundary matters.

Viral Nation's entry into Generative Engine Optimization does not prove that social engagement, creator activity or cultural visibility produces a predictable causal increase in generative citations or recommendations. Current GEO research has not established such a relationship. (2026 critical survey)

What the company does demonstrate is that the commercial definition of Generative Engine Optimization is broadening.

If people increasingly ask generative systems which products to consider, which companies are trusted, which brands are associated with a subject or what people think about a product, the information environment surrounding a brand becomes more complicated than the content published on its official website. The question begins to include how that company is represented across the larger body of information available to retrieval and generative systems.

Viral Nation is particularly relevant to that emerging problem because creator and social intelligence were part of its business before Generative Engine Optimization became a commercial category. Its AI Discovery offering connects that established social intelligence operation with the newer problem of how brands appear when generated answers become another discovery interface. (Viral Nation)

That does not establish Viral Nation's approach as the definitive future of GEO. The category is too young and the scientific evidence remains too incomplete for that conclusion. It does make the company an important participant to watch because it represents a distinct extension of Generative Engine Optimization from website optimization toward social, creator and brand intelligence.

Is GEO Becoming a Real Business and Software Category?

Generative Engine Optimization is developing beyond academic research into a recognizable market for software, agency services and organizational expertise.

Dedicated products are emerging to measure whether brands and sources appear inside generated answers, how frequently they are cited, which competitors receive greater visibility and how that visibility changes over time. Agencies are developing services around the same problem, while companies such as Viral Nation are extending the category into areas previously treated primarily as social media, creator marketing and brand intelligence. (Viral Nation)

The organizational ownership of GEO also remains unsettled. Depending on the company, responsibility could eventually sit within search, communications, content, public relations, analytics, social media or a dedicated visibility function.

That uncertainty is consistent with a category that is still forming. Generative Engine Optimization began as a research concept, developed into an academic field and is now becoming a software category, an agency service and a potential organizational function. (original GEO paper; 2026 critical survey; Viral Nation)

The participation of companies coming from different disciplines suggests that the boundaries of the category itself have not yet been settled.

Is Generative Engine Optimization Replacing Traditional SEO?

Current evidence does not establish that Generative Engine Optimization is replacing SEO. The research instead suggests that generative visibility remains dependent on many of the discovery and retrieval processes that make search optimization important in the first place. (SAGEO Arena; 2026 critical survey)

SAGEO Arena provides particularly strong evidence for this relationship because documents had to survive retrieval and reranking before they could become available for generation. The broader 2026 literature similarly identifies relevance and position as important determinants of whether information ultimately receives visibility or citation. (SAGEO Arena; 2026 critical survey)

Those findings make it difficult to separate generative visibility completely from search visibility. The interface presented to the user may be changing from a page of links toward a generated response, but the underlying system still needs methods for identifying which information deserves consideration.

A source that cannot be discovered or retrieved has little opportunity to influence the answer that follows.

Generative Engine Optimization may therefore be better understood as an expansion of the information-discovery problem rather than a replacement for everything that preceded it.

There is also evidence that the economics of traditional search are changing. Seer Interactive tracked 3,119 informational and educational queries across 42 organizations and 25.1 million organic impressions from June 2024 through September 2025. It reported that organic click-through rates for queries associated with AI Overviews declined from 1.76 percent to 0.61 percent over that period. Importantly, Seer also found that brands cited within an AI Overview had higher organic click-through rates than uncited brands within its dataset. (Seer Interactive)

That does not prove that Generative Engine Optimization is replacing SEO. It helps explain why visibility inside generated answers is becoming commercially important even while traditional search remains part of the discovery infrastructure.

What GEO Strategies Are Actually Supported by Evidence?

The strongest conclusion available in 2026 is not that publishers should follow one fixed Generative Engine Optimization formula, but that generative visibility has to be considered across the complete information pipeline.

The original GEO research demonstrated that quotations, statistics, source citations, fluency and other content characteristics can influence visibility under controlled conditions. It also demonstrated that keyword stuffing did not provide a meaningful overall benefit. (original GEO paper)

SAGEO Arena added another layer of evidence by showing that content optimization can perform differently once retrieval and reranking are included. Its results also showed that structural information can help mitigate some of the limitations produced by optimization strategies under realistic pipeline conditions. (SAGEO Arena)

The broader scientific literature adds an additional boundary. Across the 45 studies reviewed in the July 2026 critical survey, researchers did not identify a single technique that had demonstrated stable, longitudinal and cross-platform causal improvement in organic discoverability and downstream user behavior. (2026 critical survey)

Pinterest provides a useful production counterpoint. Its reported results show that substantial organic growth is possible when optimization is treated as a larger system involving semantic organization, demand prediction, multimodal understanding and internal authority rather than as a narrow collection of writing adjustments. (Pinterest GEO study)

Taken together, the evidence supports a careful approach. Publishers have reason to make information clear, relevant, technically accessible and structurally understandable while measuring how it performs across different discovery environments. They do not yet have scientific grounds to treat any short GEO checklist as a universally proven method. (original GEO paper; SAGEO Arena; 2026 critical survey; Pinterest GEO study)

What Does GEO Research Not Prove Yet?

Generative Engine Optimization research has not established that the headline percentage improvements reported in early experiments transfer directly to live production systems, nor has it established a single optimization method that reliably improves visibility across platforms and over time. (2026 critical survey)

This limitation reflects the difficulty of studying a rapidly changing class of systems whose retrieval methods, ranking processes, models and interfaces can differ substantially.

The original GEO paper established measurable effects under its experimental conditions. (original GEO paper) SAGEO Arena demonstrated that those effects need to be evaluated within a fuller retrieval environment. (SAGEO Arena) The 2026 critical survey showed that the broader literature has not yet crossed the threshold into stable, longitudinal and cross-platform causal evidence. (2026 critical survey)

Pinterest has provided valuable production evidence, but one large production implementation does not establish a universal method. (Pinterest GEO study) Viral Nation provides evidence that the commercial category is expanding into social and creator intelligence, but its entry does not establish a causal relationship between those signals and generative visibility. (Viral Nation)

The most responsible conclusion is therefore neither that Generative Engine Optimization is a marketing invention nor that GEO has already become a mature science.

It is a real field with real evidence and significant unanswered questions.

What Should Publishers and Brands Understand About GEO Now?

Publishers and brands should understand Generative Engine Optimization as part of a larger change in how information is discovered, selected and presented, while remaining cautious about claims that the field has already produced a reliable formula for generative visibility.

The evidence indicates that generated answers create a new visibility problem. Being available on the web does not guarantee retrieval. Being retrieved does not guarantee strong positioning. Strong positioning does not guarantee that a source will be cited or used in the final response. (SAGEO Arena; 2026 critical survey; Google AI Overviews study)

Research also suggests that optimizing one stage without understanding the others can produce unintended consequences. A content change that looks favorable when judged only by a generator may perform differently when retrieval and reranking are included. Structural information can affect those results, while production systems such as Pinterest's show that serious optimization can involve information architecture, semantic organization and demand modeling alongside written content. (SAGEO Arena; Pinterest GEO study)

Viral Nation introduces another dimension by bringing social and creator intelligence into the commercial Generative Engine Optimization discussion. That development does not prove that social signals directly determine generative visibility, but it demonstrates that companies are beginning to treat the problem as broader than search content alone. (Viral Nation)

The commercial GEO market will continue developing while researchers work to determine which techniques remain effective across changing systems. For publishers and brands, the useful position is therefore neither dismissal nor unquestioning adoption.

Generative Engine Optimization deserves attention because the underlying discovery environment is changing and measurable effects already exist. It deserves caution because the evidence does not yet support many of the universal claims being made around it.

The present evidence supports Generative Engine Optimization as a developing extension of the broader information-discovery problem, one that increasingly connects search, answers, retrieval, technical structure, brand representation and, through companies such as Viral Nation, the social and creator information environment.

Fact Summary

What is Generative Engine Optimization? Generative Engine Optimization is the study and practice of improving how information is discovered, selected, used and cited by systems that generate answers from retrieved sources.

Is GEO a real research field? Yes. The term was formally introduced in academic research submitted in November 2023 and accepted at ACM SIGKDD 2024. Subsequent research has expanded, tested and challenged the original findings.

Can Generative Engine Optimization improve visibility? Yes, under certain measured conditions. The original GEO research found substantial visibility changes from several optimization methods, including an approximately 41 percent relative improvement from Quotation Addition on its Position-Adjusted Word Count metric. (original GEO paper)

Does keyword stuffing work for GEO? The original research found no meaningful overall improvement from keyword stuffing and measured an approximately 10 percent decline specifically in its Perplexity evaluation. That result should not be interpreted as a universal 10 percent penalty across every generative system. (original GEO paper)

Can GEO optimization hurt visibility? Yes. SAGEO Arena found that body-text-only optimization reduced top-20 retrieval presence by approximately 9 percent, top-10 presence after reranking by approximately 16 percent and final citation by approximately 6 percent under its full-pipeline testing conditions. (SAGEO Arena)

Does schema markup matter for GEO? SAGEO Arena found that structural information could help mitigate limitations associated with optimization strategies under realistic retrieval conditions. The research does not establish schema markup as a universal shortcut to visibility. (SAGEO Arena)

Is there production evidence for Generative Engine Optimization? Yes. Pinterest researchers published a production-scale GEO case study reporting approximately 20 percent organic traffic growth from a system spanning billions of images and tens of millions of collections. (Pinterest GEO study)

Who is Viral Nation? Viral Nation is an established social media and creator marketing company that now lists AI Discovery among its services. The company says the offering is designed to help brands understand, improve and measure how AI systems describe, cite, trust and recommend them. (Viral Nation)

Has Viral Nation proved that social media activity improves GEO visibility? No. Current evidence does not establish a predictable causal relationship between social engagement or creator activity and generative citations. Viral Nation's significance is that its approach expands the commercial GEO category into social, creator and brand intelligence. (2026 critical survey; Viral Nation)

Has research identified one GEO strategy that works everywhere? No. A July 2026 critical survey of 45 studies found no reviewed technique demonstrating stable, longitudinal and cross-platform causal improvement in organic discoverability or downstream behavior. (2026 critical survey)

Is Generative Engine Optimization replacing SEO? The evidence does not establish that. Current research suggests that retrieval, relevance and position remain important to whether information becomes available for citation, making GEO closely connected to the broader search and discovery process. (SAGEO Arena; 2026 critical survey)

Sources

Aggarwal, Pranjal, Vishvak Murahari, Tanmay Rajpurohit, Ashwin Kalyan, Karthik Narasimhan, and Ameet Deshpande. "GEO: Generative Engine Optimization." arXiv:2311.09735, submitted November 16, 2023. Accepted ACM SIGKDD 2024. Original GEO paper on arXiv

Kim, Sunghwan, Wooseok Jeong, Serin Kim, Sangam Lee, and Dongha Lee. "SAGEO Arena: A Realistic Environment for Evaluating Search-Augmented Generative Engine Optimization." arXiv:2602.12187, February 2026. SAGEO Arena on arXiv

Martinez, Olivier. "Optimizing Visibility in Generative Engines: A Critical Survey of Generative Engine Optimization (2023-2026)." arXiv:2607.14035, July 2026. Critical survey on arXiv

Zhang, Faye, Qianyu Cheng, Jasmine Wan, Vishwakarma Singh, Jinfeng Rao, and Kofi Boakye. "Generative Engine Optimization: A VLM and Agent Framework for Pinterest Acquisition Growth." arXiv:2602.02961, February 2026. Pinterest GEO study on arXiv

Viral Nation. Services, including AI Discovery, and current company documentation. Viral Nation AI Discovery

Grossman, Riley, Songjiang Liu, Michael K. Chen, Mike Smith, Cristian Borcea, and Yi Chen. "How Generative AI Disrupts Search: An Empirical Study of Google Search, Gemini, and AI Overviews." arXiv:2604.27790, April 2026. Google AI Overviews study on arXiv

McDonald, Tracy. "AIO Impact on Google CTR: September 2025 Update." Seer Interactive, November 2025. Analysis of 3,119 search terms across 42 organizations, 25.1 million organic impressions. Seer Interactive CTR study