Does ranking yourself first in your own “best X” list get you recommended by AI?

Do AI answers still cite self-ranked “best X” lists?

Yes, in every AI dataset checked for this page, though Google’s AI Overviews may be cutting back. Lily Ray, an SEO consultant (founder of Algorythmic and vice-president of SEO and AI search at Amsive, per PPC Land), is critical of the tactic but writes that “they’ve undoubtedly served as one of the most effective ways to influence AI answers”.

In Peec’s sample (Peec sells AI-search monitoring), the share of citations from self-promotional listicles was about 11% overall, 3.6% on ChatGPT, 10.3% on Google AI Mode and 10.4% on Perplexity; Peec reported only those three platforms separately. Apart from AI Mode easing to 9 to 10% in February, Peec’s weekly rates held steady, and its conclusion was “No evidence of algorithmic correction over time”. That holds for December 2025 to February 2026 only. Ray’s September re-run, which has not been independently audited, found AI Overviews citing 38% fewer self-promotional lists than in June.

Two findings complicate “it gets cited”:

• ChatGPT cut listicles in general. Seer Interactive, an agency, found ChatGPT citations of any URL containing “best” or “top” fell 30% from December 2025 to January 2026 on a fixed prompt set, faster than all its citations (22.7%), while listicles in AI Overviews were flat or up. Seer notes it is one month of data.
• The citations come and go. In Ahrefs’ own experiment, in about a quarter of cases a self-promotional page was cited once for a query on an assistant and never again, and between its first and last citation a page was cited on only about one in three eligible days.

What it means. If the goal is to appear in the source list of an AI answer, a self-ranked list still does that, more on Google’s AI Mode and Perplexity than on ChatGPT in Peec’s sample. Treat each citation as a sample, not a position you hold.

If an AI cites your list, does it recommend you?

Usually not, but somewhat more often than if it had not cited your list.

Ray’s result. Across the 80 queries with an AI Overview, when a brand’s own self-promotional listicle was cited, the brand was left out of the recommendation 69% of the time: 224 of the 323 self-promotional listicles she counted as cited. She does not say whether a page cited on more than one of her three dates counts once or several times, so read 323 as cited instances, not distinct pages. By subtraction, the brand was not left out in 99 cases, or 31% (our arithmetic, not Ray’s), and Ray reports no baseline for brands without such a list. She says an established brand “may get both cited and recommended” from its own list. In a September re-run, published October 4, 2026, she found AI Overviews leaving the publishing brand out 83% of the time, up from 69%. PPC Land calls these “a consultant’s own measurements, on a small sample from one vertical and one AI surface”, with no stated method for classifying recommendations.

Scrunch’s result. Scrunch, which sells AI-visibility monitoring, analysed roughly 10,000 AI-cited URLs from May 17 to June 30, 2026, though for six of the seven assistants its sample effectively covers June 10 to 30. Holding prompt, brand and assistant fixed, the author brand’s recommendation rate went from about 4% to roughly 7% when its self-promoting listicle was cited, and a placebo test came out at about a quarter to a third of that gap. The test ran on a capped sample, roughly 6% of the eligible population. Scrunch adds that this shows what a citation does for brands already on the page, “not what adding or removing a name would do”, and that a looser comparison of a vendor’s own list against its appearances elsewhere found “a median of 0.0pp over 61 cells”. In 63.9% of responses the assistant treated the list as raw material for a balanced overview rather than adopting its number-one pick. Mentions moved more than recommendations: answers that cited a brand’s own list named the brand about 39% of the time, against about 19% when the list was not cited. Scrunch notes that this mention lift “hasn’t been through this check yet”, meaning its placebo test.

These per-assistant rates cover every answer that cited the list. Scrunch says they are not directly comparable with its 4%-to-7% figure, which comes from a narrower comparison designed to isolate the list’s effect. They are not comparable with Ray’s 69% either, because they include answers that recommend nobody.

Beyond AI Overviews. In Ahrefs’ experiment on ChatGPT, Gemini, Perplexity and Copilot, 43% of answers citing one of its conference pages never mentioned the conference brand, Ahrefs Evolve; for its tool pages, 11% failed to mention Ahrefs. That is a mention measure.

What it means. A cited list is not a vote for you. In Scrunch’s data, answers that cite your list name you about twice as often (a figure Scrunch has not yet placebo-checked), but in its controlled comparison a citation moves how often you are recommended by about three points (our subtraction), from a low base. Plan on the citation, not the crown.

Where do the “69%” and “11%” figures come from?

From two studies that measured different things, on different AI surfaces, in different months. Ray measured how often a brand is recommended once its own list is cited. Peec measured how much of what gets cited comes from self-promotional lists. Neither measured the other’s question.

The two windows do not overlap: Peec’s ends before Ray’s begins. The figures do sometimes get blended. A sponsored Search Engine Journal post restated Ray’s 224 as cases where Google “recommended a rival ranked inside it”, which is more than Ray measured. Ray’s own summary names “AI Overviews + AI Mode”, although her measured result is stated for AI Overviews. Peec’s body text calls its figure “1 in 10 citations in AI search results”, although its sample is software listicles.

What it means. When someone quotes “69%” or “11%” at you, ask which surface, which months and which denominator. The 69% is about Google AI Overviews and recommendations. The 11% is about citations in a software sample. Neither tells you what happens to your brand in ChatGPT next month.

Does your own list hand the recommendation to your competitors?

Mostly not, according to the one test of it. The open question is naming lesser-known rivals.

Ray’s evidence is worked examples, not a controlled test, such as a “best LMS for selling courses” list whose mentions of four rivals “may have helped boost them”, all four earning the AI Overview recommendation. Her September essay says that in 70% of cases the listicle “is now helping competitors get recommended while the brand that published it is left out”, with the same method caveats as above. Scrunch found that when a list is cited, a competitor is recommended instead of the author brand 24.3% of the time.

Scrunch then tested whether the list is what causes that. The placebo gap for rivals came out at +4.4 percentage points, “bigger than the real one it was meant to rule out”, and Scrunch concludes that “a brand’s listicle isn’t behind a rival being recommended”. The one case that held up was a lower-visibility brand naming other low-visibility rivals: +7 points against a placebo of -1.9. It is one of four scenarios Scrunch tested, with no correction for testing four, and Scrunch says the estimate is not precise. When a well-known brand named lesser-known rivals, the placebo reproduced about half of the rival’s lift, and Scrunch says “we can’t call this one either way”.

What it means. Scrunch’s test suggests your list is not what gets rivals recommended; it reads the rival as “likely already gaining ground for some unrelated reason”. If you name lesser-known rivals, consider naming fewer of them, or giving clear reasons when you do: for a lesser-known brand the effect held up, and for a well-known brand Scrunch could not call it either way.

When does listing yourself help?

In the one experiment built to test it, mostly when the AI did not yet know the brand and the page fitted the query. That experiment covered two brands from one company, Ahrefs, which sells the tool used, and it measured mentions, not recommendations. Ahrefs published 34 self-promotional pages (listicles, reviews and opinion pieces, not all rank-yourself-first lists) about its AI visibility tool and its conference, Ahrefs Evolve, on 5 domains, and analysed 9,886 answers from ChatGPT, Gemini, Perplexity and Copilot between February 7 and May 31, 2026.

• For a new brand, its own pages did most of the work. Of 72 new Evolve mentions in answer slots that had been empty in the February 7 to 12 baseline, 82% came in answers that cited one of Ahrefs’ pages.
• For an established brand, other people’s pages did. When Ahrefs itself broke into an empty slot, only 6% of the new mentions came from answers citing its own pages; 94% came from third-party content.
• Fit mattered. Evolve appeared in 66.4% of responses to “best SEO conferences 2026” and 15.8% for “best marketing conferences 2026”.

Scrunch’s data points partly the other way for recommendations: a well-known brand naming well-known rivals still got its own boost when its list was cited, while for a lesser-known brand naming well-known rivals the boost was too small to measure.

Ahrefs’ own advice after the test: “Add yourself to the right contexts, but don’t crown yourself”, and “Fill the gap accurately; don’t manufacture a ranking”. That is a change from December 2025, when an Ahrefs study concluded that agencies and SaaS companies “should publish self-promotional ‘best’ lists” and said there “appears to be no downside” for rankings or AI citations, while noting Ahrefs would not always put itself first.

What it means. In Ahrefs’ test, a brand the assistants did not yet know got most of its new mentions from its own pages, while an established brand got almost all of them from other people’s pages. Treat that as one company’s result, not a rule.

Did Google demote self-promotional listicles in January 2026?

At least dozens of sites lost organic traffic from around January 20, 2026, and many of them published these lists, but nobody has isolated the lists as the cause and Google confirmed no update. “May carry organic risk” is defensible; “Google punishes it” is not.

The case for a link. Ray reported on February 3, 2026, from Sistrix data, that a “well-known $8B B2B brand” lost 49% of its organic visibility between January 21 and February 2, and two other sites 43% and 29%, and called a refinement of Google’s reviews system “possible”. In June Ray wrote that “at least dozens of sites” fell around January 20, that she analysed over 40 of them, and that most used scaled AI-generated content, scaled comparison and alternative pages and, “in most cases, excessive self-promotional listicles”. She adds that many later lost visibility “across the entire domain, with drops accelerating during Google’s recent May 2026 core update”, a link to the lists that is her inference. In his August column Kevin Indig writes that he has seen the mid-January update hit listicles “anecdotally”, and calls it “one suggestive hint” of demotion that almost half of the listicles in his data come from publishers rather than brands. Publishers also won 54.0% of 4,026 direct matchups against a brand or vendor listicle on the same results page.

The case against.

• Google confirmed nothing. Its Search Status Dashboard lists the December 2025 core update (from December 11, 2025, 18 days) and then the February 2026 Discover update (February 5, 2026), with no January entry. Search Engine Land, on February 4, 2026: “Google hasn’t announced or confirmed any updates this year”. Ray herself calls it “Google’s January 2026 unconfirmed update”.
• Ray says the lists were one factor among several. In February: “only one tactic among many used by these sites”. The hit sites also used AI-generated text, artificial date refreshing and Schema.org misuse, and a few sites with heavy declines were not over-using these lists. In her own examples the lists were a small share of the blogs: 191 of about 30,000 articles, 61 of 1,420 (4%), and 10 in a blog with 1,700 indexed results. Company names are redacted, so the cases cannot be checked.
• An organic dataset found most vendor lists gaining. Indig, using data from SE Ranking (which sells SEO software), compared a fixed set of 2,400 Google queries in January and August 2026. Listicles held position one less often in August (15.1% to 15.6%, down from 18.9% to 20.2%), but he wrote: “I cannot trace it back to self-promoting listicles specifically”. 62% of vendor-authored listicles gained estimated organic traffic since January, the median page by 38%. He calls the study observational, associations not causes. “Vendor-authored” is not the same as “ranks itself first”.
• An agency that writes these pages sees no crackdown. Grow and Convert reports “no evidence of Google cracking down on self-promotional listicles” across its clients, and says Ray “never isolates and examines the rankings and traffic of just those listicles”. Its examples are self-selected, and it sells this kind of content.

Sites that lost organic traffic lost AI citations too, in one small sample: in Ray’s February 17 analysis of 11 subfolders hit in January, all 11 also lost AI search citations, 22.5% on average. Google says its AI features retrieve pages using its core Search ranking systems.

What it means. A site that leans on these pages alongside scaled AI content and scaled comparison pages is the pattern Ray saw in the sites that lost traffic in early 2026. A handful of honest self-authored comparisons on an otherwise solid site is not shown to be a problem. Watch your own data rather than the headline.

What do Google’s own documents say about self-ranked lists?

Nothing that names them. Google’s rules cover manipulation, scaled pages and review quality, and leave the question of whether your list fits to you.

• Spam policies now cover AI answers. Google’s documentation changelog for May 15, 2026 records: “Clarified that our spam policies also apply to generative AI responses in Google Search”. Neither the changelog nor the spam policy page mentions listicles or self-promotion.
• Scaled content abuse is “when many pages are generated for the primary purpose of manipulating search rankings and not helping users”. Listicles are not named.
• Inauthentic mentions. Google’s generative-AI guide, updated July 10, 2026, says seeking inauthentic “mentions” across the web “isn’t as helpful as it might seem”. It does not say whether a brand’s own list counts.
• Reviews. Google’s reviews system covers “ranked-lists of recommendations”, and on sites with a lot of review content “any content within a site might be evaluated”. Its guidance says that when you call something the best, “include why you consider it the best, with first-hand supporting evidence”, explain what sets it apart, and discuss benefits and drawbacks based on your own research. It names “a merchant who guides people between competing products” as one kind of reviewer, and nothing in it bans self-authored comparisons.

The other platforms document less. Neither OpenAI’s nor Perplexity’s help pages describe a filter for self-promotion, and Peec’s reasons for ChatGPT’s lower rate are factors that “likely explain” it, not a documented mechanism.

What it means. Google’s review guidance is the closest thing to a written standard for a vendor’s comparison page, and it asks for reasons, evidence and drawbacks. A list that ranks you first with none of those falls short of it, whoever wrote it.

Is it legal to rank your own product first?

US, EU and Swedish rules all allow comparisons with competitors on conditions, chiefly that they do not mislead. In Sweden, an unqualified “best” must be proved, and a court has said that in many cases that burden will be hard to meet. This is not legal advice, and none of these rules names self-ranked listicles.

• United States. The FTC’s comparative-advertising policy encourages naming competitors but “requires clarity, and, if necessary, disclosure to avoid deception”. Its consumer-reviews rule bars materially misrepresenting, expressly or by implication, that a website a business controls “provides independent reviews or opinions” about a category that includes its own products. The risk is implied independence, not authorship.
• Sweden. The Marketing Act requires all marketing to make clear that it is marketing and who is responsible for it (9 §), bans misleading statements about your own or another trader’s business (10 §), and allows advertising that points out a competitor only if the comparison is not misleading and objectively concerns “väsentliga, relevanta, kontrollerbara och utmärkande egenskaper” (essential, relevant, verifiable and characteristic features) (18 §). In a March 21, 2023 ruling in a case between Telia Sverige and Hi3G Access (Tre), the Patent and Market Court of Appeal said that if a “best” claim is perceived as unqualified, the advertiser must prove it, a burden that in many cases is hard to meet. Telia was barred, on pain of a SEK 2,000,000 fine, from claiming “Telia har Sveriges bästa mobilnät” (Telia has Sweden’s best mobile network) unless the marketing made clear that the claim referred to a quality test result and pointed to that test.
• EU. Article 4 of Directive 2006/114/EC permits comparative advertising only if, among other conditions, it is not misleading and “objectively compares one or more material, relevant, verifiable and representative features” (text checked on the legislation.gov.uk copy of the Directive).

What it means. None of these rules bans a vendor’s own comparison. What they target is a list that hides who is behind it or implies it is independent, and a “best” claim with no stated measure or test behind it. Saying who wrote the list, stating the criteria and resting each pick on facts a reader can check addresses those points. It does not guarantee compliance, and 18 § has further conditions.

What should you do with the “best X” pages you already have?

Keep the ones that would help a buyer even if your name were not on them, rewrite the ones that could, and retire the rest. These steps are judgement, not a tested result, drawn from Google’s review guidance, the rules above, Seer’s citation data and the vendors’ own advice.

1. List every page where you rank yourself among competitors. Include “best X”, “X alternatives” and “X vs Y” pages, and note which came from a template.
2. Check what the AI answers do with them. Ask the category question in Google AI Overviews, AI Mode, ChatGPT, Perplexity and Copilot, several times each, and in a few wordings (our guide to what decides AI recommendations explains why it takes at least seven runs per prompt, and why rewording matters more than rerunning), and record two things separately: is your page cited, and is your brand named as a pick? One run is not a measurement. Our Reddit study on measuring AI visibility covers the common sampling mistakes.
3. Check the organic trend on those pages since January 2026. In Search Console, compare the subfolder that holds them with the rest of the site. If they gained, you match the 62% of vendor-written listicles Indig saw gaining estimated traffic since January. If the subfolder fell from around January 20, look at everything else on it too, because the hit sites shared several tactics.
4. Keep a page if it states who wrote it, names the criteria, gives reasons and evidence for each pick, covers drawbacks including your own, and has its own search demand. That is close to what Google’s review guidance asks for.
5. Rewrite a page that has demand but crowns you without reasons:
   • Say at the top who wrote it and that you sell one of the products. Swedish law requires it to be clear who is behind marketing, and US rules target implied independence.
   • State the criteria, and rank on them. If you lose on a criterion, say so. “Fill the gap accurately; don’t manufacture a ranking”.
   • Give each option a reason it is best for a specific buyer, with first-hand evidence, and benefits and drawbacks from your own research.
   • Add evidence a reader can check outside your site. Seer found that the listicles growing most in ChatGPT citations, against a falling trend, included “external signals (like G2 ratings or government data) or thorough methodologies”; that measured citations, not recommendations.
   • Drop the unqualified “best”. Say best for whom, by which measure.
6. Retire pages that exist only to rank you first for a keyword, duplicate each other with a competitor’s name swapped in, or have no demand. Many pages made to rank rather than help are what Google’s scaled content abuse policy describes.
7. Put the effort where established brands get named: other people’s pages. For an established brand, 94% of new AI mentions in Ahrefs’ test came from third-party content. In Ray’s reading, which she says is hard to confirm, the recommended brands were the ones talked about and linked to elsewhere. Do not buy mentions: Google says inauthentic ones are not as helpful as they might seem.
8. Measure again in a month, and keep the raw answers. Citations from these pages are unstable; in one practitioner’s 1,000-prompt test, 11.6% of ChatGPT prompts switched primary search source between runs; and Ray’s own figures moved between June and September.

What it means. No dataset here tests it directly, but the pages least exposed to every risk on this page are the ones a buyer would trust knowing who wrote them. Make the authorship obvious and the criteria checkable.

What can’t this evidence tell you?

• No baseline in the headline study. Ray reports how often self-promoters are left out, not how often brands without a list are left out. Without that, 69% cannot be read as an effect of the list. Her coded results behind 224 of 323 are not published, so the 69% cannot be re-counted.
• No test of honest comparisons. No source checked tests whether a disclosed, criteria-led vendor comparison earns more recommendations than a self-crowning one, so the keep, rewrite or retire advice is inference.
• Vendor interest everywhere. Peec, Ahrefs and Scrunch sell AI-visibility or SEO tools; Seer, Grow and Convert and Ray sell agency or consulting services, and Ray’s sample includes sites she has worked with since the January drops. Neither Ray’s September figures nor Scrunch’s study has been independently audited.
• Software only. Ray’s and Peec’s samples are software queries. Brand and vendor listicles appeared in the top 10 on 46.2% of B2B software results pages in Indig’s data, against 3.3% in food and 6.2% in travel, so the pattern may not carry over to law firms, clinics or local services.
• Units differ. Ray counts cited instances across three dates; Peec counts citations in a top-1,000 weekly sample; Scrunch counts answers; Ahrefs counts mentions. Do not compare the numbers across studies.

FAQ

Do AI assistants cite “best X” lists where the publisher ranks itself first?

Yes. In Peec’s study of 232,000 citations across 13,000 software listicles from December 2025 to February 2026, about 11% came from self-promotional listicles (3.6% on ChatGPT), with no sign that the platforms got better at filtering them out over the 12 weeks. Peec sells AI-search monitoring, and the figure is a share of its software sample, not of all AI citations.

If an AI answer cites my own “best” list, will it recommend my brand?

Usually not. In Lily Ray’s study of Google AI Overviews for B2B software queries, April to June 2026, the brand behind a cited self-promotional listicle was left out of the recommendation in 224 of 323 cases, or 69%. Scrunch, across seven AI assistants in May and June 2026, found a brand’s recommendation rate rose only from about 4% to roughly 7% when its own list was cited.

Did Google penalise self-promotional listicles in January 2026?

It is not established. Lily Ray reported at least dozens of sites losing organic traffic from around January 20, 2026, but says self-promotional listicles were one of several potential issues. Google confirmed no January update, and Kevin Indig found 62% of vendor-written listicles (not necessarily self-ranked) gained estimated Google traffic between January and August 2026.

Is it legal to rank your own product first in a “best” list?

US, EU and Swedish rules all allow comparisons with competitors on conditions, chiefly that they do not mislead. In Sweden, an unqualified “best” must be proved, and a court has said that in many cases that burden will be hard to meet. A list that says who wrote it, states its criteria and compares verifiable features addresses what these rules focus on. This is not legal advice.

Should I delete my self-ranked “best X” pages?

Not all of them. Keep pages that state who wrote them, name their criteria and give reasons, evidence and drawbacks for each option, which is what Google’s review guidance asks for. Rewrite pages that have search demand but crown you without reasons. Retire near-identical pages made only to rank you first, the pattern Google’s scaled content abuse policy describes. No study has yet tested whether a rewritten, disclosed comparison earns more AI recommendations.

Related pages on this site

• What decides whether an AI recommends you, and what to do about it: the four gates between a question and a recommendation, and why third-party sources decide whether you are named.
• How to write content that LLMs retrieve and cite: every common AEO practice rated by its evidence.
• How AI agents choose products for consumers: what moves an agent’s pick, including ratings, badges and third-party reviews.
• Where AI search goes next: what the platforms sell, measure and disclose.
• “Useful diagnostic, terrible KPI”: what 200,753 Reddit posts say about measuring AI visibility: how to sample AI answers without fooling yourself.

Sources

All sources read at source on October 24, 2026. Sources from companies that sell AI-visibility, SEO or monitoring products, or agency and consulting services around them, are marked (vendor), (agency) or (consultant). Pages that blocked automated access were read from Wayback Machine captures of the original URL, as noted.

The headline studies

• Lily Ray, “Why Calling Yourself the ‘Best’ Could Be Helping Your Competitors Win in AI Search”, Substack, June 17, 2026 (consultant). Accessed October 24, 2026.
• Peec AI (Tom Wells), “Self-promotional listicles analysis: Data from 232,000 citations”, March 3, 2026 (vendor). Accessed October 24, 2026.
• Lily Ray, “Prediction: the next massive Google update is around the corner”, Substack, October 4, 2026 (consultant). Accessed October 24, 2026.
• Scrunch (Niharika Sharma), “What happens when you crown yourself #1: Inside thousands of listicles and their impact on AI answers”, September 24, 2026 (vendor). Accessed October 24, 2026.
• Ahrefs (Mateusz Makosiewicz), “Self-Promotional Content Works—Until It Backfires (AI SEO Experiment)”, July 6, 2026 (vendor). Accessed October 24, 2026.
• Ahrefs (Glen Allsopp), “Do Self-Promotional ‘Best’ Lists Boost ChatGPT Visibility? Study of 26,283 Source URLs”, December 4, 2025 (vendor). Accessed October 24, 2026.
• Seer Interactive, “The Listicle Window Is Closing in AI Search: 30% Decline MoM”, February 25, 2026 (agency). Accessed October 24, 2026.

Organic search

• Lily Ray, “Is Google Finally Cracking Down on Self-Promotional Listicles?”, Substack, February 3, 2026 (consultant). Accessed October 24, 2026.
• Lily Ray, “Are Citations in AI Search Affected by Google Organic Visibility Changes?”, Substack, February 17, 2026 (consultant). Accessed October 24, 2026.
• Kevin Indig, “Has Google demoted listicles?”, Search Engine Land, August 26, 2026, SE Ranking data (read from a Wayback Machine capture). Accessed October 24, 2026.
• SE Ranking, homepage (“AI SEO Software”), cited only to label the data provider in Indig’s column (vendor). Accessed October 24, 2026.
• Danny Goodwin, “Google may be cracking down on self-promotional ‘best of’ listicles”, Search Engine Land, February 4, 2026 (read from a Wayback Machine capture). Accessed October 24, 2026.
• Devesh Khanal, “Self-Promotional Listicles Aren’t Bad For SEO. Bad Content Is”, Grow and Convert, May 19, 2026 (agency). Accessed October 24, 2026.
• Google Search Status Dashboard, “History for Ranking”. Accessed October 24, 2026.

Google, OpenAI and Perplexity documentation

• Google Search Central, “Latest Google Search Documentation Updates”, May 2026 entries. Accessed October 24, 2026.
• Google Search Central, “Spam policies for Google web search”, updated August 28, 2026. Accessed October 24, 2026.
• Google Search Central, “Optimizing your website for generative AI features on Google Search”, updated July 10, 2026. Accessed October 24, 2026.
• Google Search Central, “Write high quality reviews”, updated December 10, 2025. Accessed October 24, 2026.
• Google Search Central, “Google Search’s reviews system and your website”, updated December 10, 2025. Accessed October 24, 2026.
• OpenAI Help Center, “Searching the web with ChatGPT” (read from a Wayback Machine capture). Accessed October 24, 2026.
• Perplexity Help Center, “How does Perplexity work?”. Accessed October 24, 2026.
• Danny Goodwin, “ChatGPT citations change when hidden search pipelines switch”, Search Engine Land, July 8, 2026 (read from a Wayback Machine capture). Accessed October 24, 2026.

Press coverage

• Luis Rijo, “Google AI Overviews cut self-ranking listicle citations 38%, Lily Ray finds”, PPC Land, October 5, 2026. Accessed October 24, 2026.
• Tautvydas Vasiliauskas, “AI Search: Is Your Content Strategy Accidentally Recommending Your Competitors?”, Search Engine Journal, July 7, 2026 (sponsored by FirstPromoter; cited only as an example of how the figures get restated). Accessed October 24, 2026.

Advertising law

• 16 CFR § 465.6, Trade Regulation Rule on the Use of Consumer Reviews and Testimonials, eCFR. Accessed October 24, 2026.
• 16 CFR § 14.15, FTC Statement of Policy Regarding Comparative Advertising, eCFR. Accessed October 24, 2026.
• Marknadsföringslag (2008:486), Riksdagen; consolidated text also checked at lagen.nu. Accessed October 24, 2026.
• Svea hovrätt, Patent- och marknadsöverdomstolen, judgment of March 21, 2023, case PMT 14821-21 (Telia Sverige AB and HI3G Access AB). Accessed October 24, 2026.
• Directive 2006/114/EC, Article 4, EU-origin copy on legislation.gov.uk. Accessed October 24, 2026.

How to cite: Copy for Originals, “Does ranking yourself first in your own ‘best X’ list get you recommended by AI?”, October 24, 2026, https://tools.copyfororiginals.com/self-promotional-listicles-ai-search/.