By Sagoh 
In the rapidly evolving landscape of search engine optimization (SEO), a counterintuitive and highly disruptive trend has emerged. For years, business-to-business (B2B) software companies and software-as-a-service (SaaS) providers have relied on self-promotional "best-of" listicles to dominate search engine results pages (SERPs). By publishing articles titled "Best [Category] Software" and ranking their own product at the top, brands successfully captured high-intent transactional traffic.
However, a groundbreaking study by prominent search analyst Lily Ray reveals that Google’s generative AI search feature, AI Overviews (formerly known as the Search Generative Experience, or SGE), is turning this strategy on its head.
The analysis reveals a stark reality: Google is increasingly using these self-serving listicles as source material to train and inform its AI-generated answers, while systematically excluding the authoring brands from the actual product recommendations. Instead, the search engine recommends the author’s direct competitors.
1. Main Facts: The "Citation vs. Recommendation" Gap
The core finding of Ray’s research is a phenomenon that digital marketers are calling the "citation paradox." While B2B brands have historically used biased listicles to influence search algorithms and prospective buyers, Google’s AI Overviews are leveraging the structured data within these articles to formulate AI responses, only to recommend better-known competitors.
The study analyzed B2B software queries and revealed that in 69% of analyzed cases, Google’s AI Overviews cited a brand’s self-promotional "best" listicle as a reference source while completely excluding that brand from the recommended list of software providers.
+-------------------------------------------------------------------+
| THE CITATION PARADOX IN NUMBERS |
+-------------------------------------------------------------------+
| 69% of cases: Brand's listicle is cited, but brand is excluded |
| from the actual recommendations. |
| |
| 30% - 50% drop: Organic visibility loss for brands relying |
| heavily on self-promotional "best-of" pages. |
+-------------------------------------------------------------------+This divergence highlights a critical paradigm shift in search engine behavior: a citation is no longer synonymous with a recommendation.
For B2B brands, this represents an existential threat to content marketing ROI. Companies are essentially funding the creation of high-quality, research-intensive content that Google uses to divert potential customers directly into the funnels of their market rivals.
2. Chronology of the Shift: From SGE to AI Overviews
To understand how this dynamic materialized, it is necessary to trace the timeline of Google’s integration of generative AI into its core search product, alongside concurrent algorithmic updates.
2023 (May) 2024 (April-June) 2024 (May) 2024 (Mid-Year)
│ │ │ │
├──────────────────────┼─────────────────────────┼─────────────────────────┤
│ │ │ │
Google launches Lily Ray conducts Google rolls out Marked decline in
SGE in Beta; longitudinal study AI Overviews to organic traffic for
testing AI-driven tracking 100 B2B US public; algorithm self-ranking SaaS
summaries. queries over 3 dates. prioritizes UGC/Reddit. listicles observed.The SGE Beta Era (May 2023 – April 2024)
Google introduced the Search Generative Experience (SGE) as an opt-in experiment in Search Labs. During this phase, SEO professionals observed that the AI frequently pulled information directly from top-ranking organic pages, often mirroring the organic SERPs. Brands publishing "best [category] software" lists enjoyed high visibility, as their pages were both ranking organically and serving as the primary sources for SGE snapshots.
The Spring Transition and Lily Ray’s Study (April – June 2024)
As Google prepared for a wider public rollout, the underlying algorithms governing source selection and entity evaluation underwent significant adjustments. Lily Ray initiated a longitudinal study to monitor these shifts, capturing data across three critical checkpoints:
- April 15, 2024: Baseline tracking during the late experimental phase of SGE.
- May 15, 2024: Mid-point analysis, coinciding with Google’s official announcement at the I/O conference that SGE was transitioning to "AI Overviews" and rolling out to hundreds of millions of users in the United States.
- June 8, 2024: Post-rollout assessment to measure the stabilized behavior of the live AI Overviews engine.
The Post-Rollout Reality (June 2024 – Present)
Following the public launch, the organic visibility of traditional self-promotional listicles began to plummet. Earlier research conducted by Ray and reported by Search Engine Land indicated that several prominent SaaS and B2B brands suffered immediate 30% to 50% declines in organic visibility on pages that relied heavily on self-ranked "best-of" content.
This drop occurred because Google began actively penalizing biased, self-serving comparisons under its Helpful Content System and core algorithmic updates, while simultaneously utilizing those same pages to feed its AI Overview summaries.
3. Supporting Data: Inside the Numbers of the B2B Query Analysis
The empirical backbone of this shift is detailed in Ray’s analysis, which utilized advanced rank-tracking and data-mining tools to dissect how Google processes B2B software queries.
Methodology
Using Ahrefs Brand Radar, Ray tracked and collected data for 100 highly competitive B2B queries structured around the format: "best [category] software" (e.g., best CRM software, best project management software, best email marketing tools). At each of the three checkpoints (April 15, May 15, and June 8), the research evaluated:
- The text generated within the AI Overview.
- The specific URLs cited in the link cards accompanying the AI Overview.
- The actual brand entities recommended to the user within the AI-generated text.
Key Data Findings
The 69% Exclusion Rate
The most alarming statistic for B2B marketers is the 69% exclusion rate. In nearly seven out of ten instances where a brand’s self-promotional listicle was selected by Google as a cited source, the brand itself was omitted from the AI’s list of recommended software solutions. The AI algorithm successfully extracted the industry context, definitions, and criteria from the brand’s article, but filtered out the authoring brand in favor of external market leaders.
The Triumph of Established Brand Entities
The data showed that Google’s AI did not distribute recommendations democratically. Instead, the AI Overviews consistently favored:
- Category Leaders: Brands with dominant market share and high volume of branded search queries.
- Widely Mentioned Third-Party Entities: Software solutions frequently cited across independent, authoritative domains.
- Strong Link Profiles: Brands possessing a high volume of editorial backlinks from reputable news outlets, educational institutions, and industry journals.
The Rise of User-Generated Content (UGC) and Review Aggregators
The study documented a dramatic realignment in the types of domains cited by Google for "best" software queries. While brand-owned blogs lost ground, two categories saw unprecedented growth:
- Review Platforms: G2, Capterra, TrustRadius, and Gartner Peer Insights maintained or increased their dominance as trusted sources.
- User-Generated Content (UGC): Citations of Reddit and Quora threads surged. Google’s algorithms increasingly treated discussions on Reddit as authentic, unbiased user experiences, prioritizing them over polished corporate blog posts.
4. Google’s Search Philosophy and Official Responses
To understand why this divergence occurs, it is necessary to examine Google’s broader search quality guidelines, its architectural approach to Information Retrieval (IR), and its public statements regarding AI-generated search results.
The Objective of AI Overviews
Google’s stated goal for AI Overviews is to provide users with a synthesized, objective overview of a topic, pulling from the high-quality web index to save searchers the time it would take to click through multiple websites.
When a user searches for the "best" software in a category, Google’s AI is programmed to identify consensus. If five different sources recommend Software A, and only one self-published source recommends Software B (which happens to be the publisher of that source), the AI’s natural language processing (NLP) models identify the self-recommendation as an outlier or a biased claim.
The Role of E-E-A-T and Search Quality Rater Guidelines
Google’s Search Quality Rater Guidelines place heavy emphasis on E-E-A-T (Experience, Expertise, Authoritativeness, and Trustworthiness).
┌──────────────────────┐
│ TRUSTWORTHINESS │
│ (The Central Pillar) │
└──────────┬───────────┘
│
┌────────────────────────┼────────────────────────┐
▼ ▼ ▼
┌─────────────────┐ ┌─────────────────┐ ┌─────────────────┐
│ EXPERIENCE │ │ EXPERTISE │ │ AUTHORITATIVENESS│
│ Real-world use │ │ Credentials & │ │ Industry brand │
│ & UGC content │ │ deep knowledge │ │ equity & links │
└─────────────────┘ └─────────────────┘ └─────────────────┘Within this framework, Trustworthiness is considered the most critical pillar. Under Google’s guidelines:
- An article written by a company claiming its own product is the "best" inherently lacks objectivity.
- The system is designed to distinguish between independent editorial reviews and commercial self-promotion.
- Consequently, while the information within a brand’s listicle may be structured well enough to be crawled and cited as a definition or a list of options, the source itself is not deemed trustworthy enough to receive an endorsement from the AI.
Google’s Official Stance on Citations
When questioned about the impact of AI Overviews on web traffic and publisher visibility, Google has consistently maintained that AI Overviews are designed to drive high-value traffic to the web. Google executives have asserted that the link cards within AI Overviews receive higher click-through rates than traditional organic listings because they appear in a highly context-rich environment.
However, Google has not directly addressed the competitive disadvantage of the "citation paradox," where a brand’s content is used to validate and direct traffic to its direct competitors.
5. Strategic Implications for B2B SaaS and SEO
The findings of Lily Ray’s study signal the end of an era for traditional B2B content marketing. The "double-dip" strategy—where a brand ranks both its product landing page and its self-made comparison listicle on the first page of Google—is no longer viable. Marketers must completely re-engineer their search strategies to survive in an AI-first environment.
The Redefinition of SEO ROI: "A Citation is Not a Recommendation"
Marketers must separate "visibility" from "customer acquisition." Appearing in the citation cards of an AI Overview may look promising in an SEO report, but if the text of the AI Overview advises the user to buy a competitor’s product, the citation is functionally counterproductive.
+-------------------------------------------------------------------------+
| THE SAAS MARKETING RE-ALIGNMENT |
+-------------------------------------------------------------------------+
| OLD PLAYBOOK: |
| Write a "Best [My Category] Software" blog post. |
| Put your own brand at #1. |
| Build backlinks to the post to rank organically. |
| |
| NEW PLAYBOOK: |
| Optimize external brand mentions across Reddit, G2, and Capterra. |
| Build brand authority through Digital PR and industry partnerships. |
| Pivot internal content to deep, unbiased, and original research. |
+-------------------------------------------------------------------------+Actionable Strategies for B2B Brands
To navigate this new paradigm, B2B and SaaS brands should implement the following strategic pivots:
1. Shift from Self-Promotion to Objective Analysis
If brands continue to publish comparison content, they must adopt a highly objective, journalistic tone. Instead of declaring themselves the absolute "best," content should focus on specific use cases (e.g., "Our software is best for enterprise companies needing deep API integrations, whereas Competitor A is better suited for small teams").
This level of objectivity aligns with Google’s search quality guidelines and reduces the likelihood of the content being flagged as purely self-serving.
2. Invest Heavily in Digital PR and Off-Page Entity Building
Because Google’s AI relies on web-wide consensus to determine which brands to recommend, off-page SEO and digital PR are more critical than ever. Brands must ensure they are mentioned positively on:
- Independent industry news sites.
- Reputable technology blogs.
- Major business publications (e.g., Forbes, TechCrunch, VentureBeat).
- Academic and industry whitepapers.
The goal is to build a robust "entity footprint" that proves to Google’s knowledge graph that the brand is a legitimate market leader.
3. Dominate Third-Party Review Ecosystems
Since Google AI Overviews heavily cite and recommend based on data from review aggregators like G2, Capterra, and TrustRadius, B2B companies must prioritize review acquisition campaigns. A steady stream of authentic, highly-rated reviews on these platforms will directly influence the AI’s propensity to recommend the brand.
4. Optimize for User-Generated Content (UGC) and Forums
With Reddit and Quora experiencing a massive surge in AI Overview citations, brands must actively monitor and participate in these communities. This does not mean spamming forums with promotional links, which leads to community bans. Instead, it requires:
- Engaging in genuine discussions.
- Answering technical user questions.
- Encouraging satisfied customers to share their real-world experiences on subreddits dedicated to the industry.
5. Diversify Traffic Acquisition Channels
Relying solely on informational B2B search queries is increasingly risky. Brands must diversify their marketing mix by investing in:
- Direct-to-consumer brand building: Ensuring users search for the brand name directly, bypassing generic "best software" queries entirely.
- Email and community marketing: Building owned audiences that cannot be disrupted by algorithmic changes.
- Video and visual search: Optimizing for platforms like YouTube, which are increasingly integrated into search experiences but operate on different algorithmic parameters.
Conclusion
The data compiled by Lily Ray serves as a stark warning to the B2B SaaS sector: the search landscape has fundamentally changed. As Google AI Overviews continue to mature, the algorithmic systems that power them will only grow more sophisticated at identifying and filtering out self-promotional bias.
Brands that adapt by focusing on genuine authority, third-party validation, and objective content will emerge as the recommended leaders in the AI-driven search era. Those that cling to the outdated playbook of self-ranking listicles risk watching their own content serve as the stepping stone for their competitors’ success.
Nana 
Jia Lissa 
Evan Lee Salim 