How Can Brands Adapt to Google's Reduced Citations of Self-Promotional Content?

What changed in Google AI Overviews?

Google AI Overviews appear to be citing fewer vendor-authored pages that rank their own product first. PPC Land reported on October 5, 2026 that Lily Ray measured 100 best B2B software queries in September and found a 38% decline in citations of these self-ranking listicles compared with her June measurement. The same report says the page author was left out of the final recommendation in 83% of AI answers that cited the page.

That result does not establish a new published Google rule, nor does it prove that every first-party comparison page is suppressed. It is a small, practitioner-led study in one vertical and one search surface. Its practical meaning is still hard to ignore: a brand can supply information to an AI answer without being the brand that answer recommends.

Google's public guidance supports a more cautious reading. Google says AI Overviews surface relevant links and may use query fan-out, which runs related searches across subtopics and data sources. A single self-promotional page is therefore competing with a wider evidence set, not simply trying to rank for one familiar query.

The change matters because self-ranking pages are designed around an obvious conflict. A company can explain its own strengths, but a page that presents itself as neutral while always placing its own product first gives a reader little reason to trust its selection method. AI Overviews do not need to accept that framing when other sources can support, challenge, or replace it.

1. Treat the 38% figure as a directional finding from a dated study, not as a universal Google benchmark.
2. Separate citation from recommendation. A page may be cited for a fact while another source or brand receives the recommendation.
3. Assume AI Overviews can evaluate a category through multiple related queries and sources.

Who does reduced citation of self-promotional content affect?

The immediate risk is highest for B2B software companies that publish high-volume best-of, alternatives, and comparison pages in which their own product consistently wins. These pages may still attract classic search traffic, but the cited source and the recommended brand can now diverge in AI Overviews.

It also affects agencies and publishers that sell paid placements in category pages. A brand mention bought to influence an AI answer has two weaknesses: it may not be cited, and it may place the buyer next to claims it cannot substantiate. Google says its systems prioritize helpful, reliable information created to benefit people rather than content created to manipulate search rankings.

The risk is not limited to software. Local and service businesses can create the same credibility problem when a page calls itself a guide to the best provider in a city but has no transparent method, local evidence, or meaningful discussion of alternatives. A generic template repeated across locations makes the problem worse because it adds coverage without adding proof.

This does not mean brands should stop publishing comparison content. Buyers need comparisons. The distinction is whether the page helps someone make a decision or simply stages an outcome the publisher has already chosen. A fair comparison can explain where the publisher fits, where it does not, and what a reader should verify before buying.

1. B2B SaaS teams with self-ranking category pages face the most direct exposure.
2. Placement programs that trade on claimed AI visibility face an evidence and trust problem.
3. Local businesses should audit city pages that rely on copied claims instead of local proof.
4. Editorial teams should protect legitimate comparison content by showing their method and scope.

What does the before-and-after evidence show?

The dated evidence points to a deteriorating outcome for the publisher of a self-ranking listicle. In June 2026, Ray reported that brands ranking themselves first were excluded from the actual recommendation 69% of the time. Her September 2026 measurement reported exclusion in 83% of AI answers that cited those pages, alongside 38% fewer citations of the pages themselves.

The figures are useful because they specify both the surface and the limitation. They concern Google AI Overviews, 100 best B2B software queries, and one consultant's methodology. They do not measure every category, every country, classic organic results, ChatGPT, Perplexity, Gemini, or Google AI Mode. Do not turn them into a blanket claim about AI search.

The action is clearer than the causal explanation. If your comparison page can be cited while your brand is excluded, publishing more versions of the same page is not a defensible response. Audit whether the page contains independently checkable facts, a selection method, meaningful trade-offs, and information a buyer cannot get from your product marketing page.

Google's own documentation says pages eligible as supporting links in AI Overviews must be indexed and eligible to appear with a snippet in Google Search. It also says there are no additional technical requirements for AI features. The work is therefore not a special markup trick. The work is making the page useful, accessible, and trustworthy enough to deserve inclusion.

1. Before, June 2026: the self-ranking brand was omitted from the recommendation 69% of the time in Ray's reported measurement.
2. After, September 2026: the omission rate was 83%, and citations of vendor-written best-of pages were 38% lower than in June.
3. What to do: replace outcome-led listicles with buyer-led comparison pages supported by a stated method and verifiable evidence.

How should brands respond without abandoning comparison content?

Brands should publish comparisons that make the reader's decision easier even if the reader does not choose them. Start by defining who the comparison is for, the use case being evaluated, the information date, and the criteria used. Then support factual claims with product documentation, pricing pages, first-party research, or clearly attributed customer evidence.

A credible first-party page can still argue for the brand. It should do so after it has established the relevant trade-offs. For example, a page can state that a product is a better fit for a defined workflow, team size, integration need, or service model. It should not claim to be the best option for every buyer without explaining how that conclusion was reached.

Use original material where it can be checked. That might include a transparent has matrix, dated pricing inputs, a documented test process, technical implementation notes, or original research with a stated sample and method. Do not manufacture an editorial voice through fictional authors, invented customer stories, or unsupported performance claims.

Google's people-first guidance tells publishers to assess whether content is helpful and reliable, and to look at the pages most affected when diagnosing a drop. That is a sensible audit process for AI Overviews too. Review the page inventory by query intent, identify where self-promotion substitutes for evidence, and rebuild the highest-value pages first.

For a broader operating view, treat the result as an AI search visibility problem rather than a single-page SEO problem. Track which questions matter, which sources AI answers cite, whether your brand appears in the recommendation, and how that compares with named competitors. That is closer to Citation Share than a count of pages published.

1. State the audience, comparison scope, criteria, and date of the review.
2. Cite product facts to their primary documentation and label judgments as judgments.
3. Show real trade-offs, including situations where a competitor may fit better.
4. Remove unsupported superlatives, fictional authors, and copied category claims.
5. Measure answer presence and citation patterns across the questions buyers actually ask.

What should a self-ranking listicle audit include?

An effective audit begins with the pages most likely to shape buyer recommendations: best-of pages, alternatives pages, versus pages, category guides, and local service lists. Record the query each page targets, whether it ranks the publishing brand first, the stated evidence for each inclusion, and the page's last substantive review date.

Next, inspect the main content rather than only the title tag or schema. Google says people-first content should benefit people, and its documentation advises publishers to look closely at the pages and search types affected by drops. A useful audit asks whether the page gives a buyer enough information to compare options without accepting the publisher's conclusion on trust.

Then decide whether each page should be retain, rebuild, consolidate, or remove. Retain pages with a real method and current evidence. Rebuild pages that contain useful subject knowledge but use unsupported rankings. Consolidate near-duplicates that repeat the same conclusion across thin variants. Remove pages that cannot be made accurate without inventing evidence.

Do not respond by adding a special file or a new type of markup. Google explicitly says there is no special schema.org structured data and no new machine-readable file needed to appear in AI features. Technical hygiene still matters: the page must be indexable, eligible for a snippet, and easy for Google to crawl and understand.

1. Map every self-ranking page to its target question and business purpose.
2. Check the selection method, source trail, update date, and unsupported claims.
3. Prioritize pages that influence commercial decisions or repeat across many templates.
4. Classify each page as retain, rebuild, consolidate, or remove.
5. Confirm indexability and snippet eligibility before treating content quality as the only issue.

How should brands measure the response?

Measure the answer, not just the page. Search Console includes AI has traffic within its Web search reporting, according to Google, and Analytics can show what visitors do after they arrive. Those reports help diagnose traffic, but they do not independently tell a brand whether AI Overviews recommend it for its category questions.

Build a prompt set around the commercial questions that matter: category selection, alternatives, comparisons, implementation concerns, and local intent where relevant. Capture whether an AI answer appears, which domains it cites, whether your brand is named, and whether a competitor receives the recommendation. Keep dates and wording because AI responses and their link sets can vary.

Use that evidence to decide what content deserves editorial investment. If a page repeatedly supplies context but a competitor receives the recommendation, examine the gap. It may be missing a clear proof point, a relevant use case, transparent comparison criteria, or fresh information. Do not assume that adding more praise for your own brand will solve it.

The durable goal is not to force a model to repeat a preferred line. It is to create accurate, current, useful material that can stand up beside independent sources. That is the practical discipline behind AI search visibility, and it is a more stable response than trying to game a changing citation pattern.

1. Track AI Overview appearance, cited domains, brand recommendation, competitor recommendation, and date.
2. Use Search Console and Analytics for traffic and post-click behavior, then pair them with answer-level observation.
3. Refresh evidence when products, pricing, policies, or the underlying question changes.
4. Judge success by improved presence in relevant answers, not by the number of self-promotional pages published.

Sources

Source: PPC Land

Lily Ray's September 2026 study of 100 best B2B software queries reported 38% fewer citations of vendor-written self-ranking best-of pages than in June, and omission of the publishing brand from 83% of recommendations that cited those pages. PPC Land, 2026-10-05

Source: Google Search Central

Google explains that AI Overviews may use query fan-out across related searches and that standard SEO best practices apply. Supporting-link eligibility requires indexing and eligibility for a Google Search snippet, with no additional technical requirements. Google Search Central, 2025-12-10

Source: Google Search Central

Google says its automated ranking systems prioritize helpful, reliable information created to benefit people rather than content created to manipulate search rankings, and advises reviewing affected pages and search types when diagnosing drops. Google Search Central, 2025-12-10