--- title: "Has Google Demoted Listicles? What a 60,000-Query Study Says - AstroDev" description: "Google demoted listicles at the top of results in 2026, yet AI engines still cite them most. Where the roundup tactic works, and where it collapses." url: "https://astrodev.carlosarias.com/blog/reviews/are-listicles-still-worth-writing-seo" --- [Reviews](/categories/reviews) # Has Google Demoted Listicles? What a 60,000-Query Study Says Google demoted listicles at the top of results in 2026, yet AI engines still cite them most. Where the roundup tactic works, and where it collapses. [Carlos Arias](/authors/carlos-arias) · September 28, 2026 · 5 min read ![A ranked list of glass panels, the top row dimming as an AI answer box rises beside it.](/_astro/cover.C67i2h2x_11NbgX.webp) *A ranked list of glass panels, the top row dimming as an AI answer box rises beside it. AI-generated illustration by Carlos Arias .* On-brand editorial cover for an article titled "Has Google Demoted Listicles? What a 60,000-Query Study Says". Sophisticated, minimal conceptual illustration on a very dark ink background (#111318) with a single restrained warm accent glow. High-end business-publication aesthetic, subtle depth, cinematic soft light. No text, no words, no letters, no logos, no UI labels. Prompt sent to Higgsfield · nano_banana_pro · 3:2 Has Google demoted listicles? Partly, and the nuance is the whole story. The data shows Google demoted listicles at the very top of its results in early 2026, while AI engines moved the other way and now cite the format more than any other. So before you mass-produce roundup pages to farm AI citations, know exactly which half of that split you are betting on. This is a review of what a 60,000-query study and three independent analyses found, and where the “self-ranking roundup” tactic still pays versus where it is quietly failing. ## The short answer: a trim at the top, not a purge Listicles lost ground at position one and held it almost everywhere else. In Kevin Indig’s August 2026 analysis for Growth Memo, listicles at rank one fell from a January share of 18.9% to 20.2% down to 15.1% to 15.6% by August. Across the top three, the January band of 35.9% to 39.0% eased to roughly 32.0% to 33.0% (Growth Memo, August 2026). That is a real dip. It is not a collapse. Read against traffic, the picture gets stranger. Vendor listicles actually gained a 38% median traffic lift in the months after Google’s January adjustment, because the demotion hit self-serving pages hardest and left genuinely useful roundups room to rise (PPC Land, 2026). The format is not dead. The lazy version of it is losing. ## What the 60,000-query study actually measured Scale is why this study is worth citing at all. The analysis covered 60,000 US-English desktop Google queries across 15 verticals, pulling 5.32 million organic-result rows during a 47-hour window in August 2026 (Search Engine Land, 2026). A hunch drawn from a dozen SERPs is folklore. A controlled read across millions of rows is evidence. The same dataset drives our companion piece on listicle ranking signals, where the durable finding was freshness, not format. A stale or unreadable date carried 56% lower odds of a top-3 slot after controls. That matters here for one reason. When people say Google punished listicles, the mechanism was rarely “it is a list.” It was thin, dated, self-interested content that happened to wear a list format. ## Where roundup pages still win: AI citations Here is the side of the split that keeps the tactic alive. Listicles are the single most-cited page type inside ChatGPT answers, accounting for 43.8% of 26,283 source URLs the study examined (Growth Memo, August 2026). Nearly half of what ChatGPT pulls from is a list. That includes pages where a brand ranks itself first, because across ChatGPT, Gemini, Perplexity, Copilot, and Google’s own AI surfaces, the study found no citation penalty for self-promotion. So a well-built roundup remains one of the most reliable formats for listicle citations in ChatGPT and its peers. The catch is that the ground is moving underneath it. ChatGPT cut its listicle-citation rate by roughly 30% between December 2025 and January 2026, and several SaaS brands lost between 29% and 49% of organic visibility within weeks of Google’s January change (Growth Memo, August 2026). Being cited is not the same as being safe. It is a position you have to keep re-earning as engines retune what they trust. ## Where the self-serving listicle tactic is collapsing Now the failure mode. A citation is not a recommendation, and conflating the two is the mistake burning marketing teams right now. Lily Ray analyzed 100 B2B “best [category]” queries in Google’s AI Overviews, checking each three times between April and June 2026. Of the 80 prompts that triggered an AI Overview, self-promotional listicles were cited 323 times. In 224 of those cases, Google cited a brand’s own “best” page but then recommended a competitor instead (Search Engine Land, 2026). That is a 69% competitor-recommendation rate on your own content. You wrote the list that named yourself number one. Google quoted it, then told the buyer to look at someone else. The engine treats a self-ranking page as a source of options, not as a verdict it should trust. The more transparently self-serving the page, the more it reads as raw material for the model to reorder in someone else’s favor. This is the self-serving listicles demotion in its real form. It is not a format penalty. It is a trust penalty, applied to pages that exist to rank the publisher rather than help the reader, and it shows up as your own page handing the sale to a rival. ## The verdict: build for the reader, not the scheme If your plan is to spin up roundup pages for AI search at volume, each engineered to seat your product at the top, this study says the return is shrinking and the downside is concrete. You risk citing yourself into a competitor’s win. The tactic still works under narrow conditions: - Genuine comparison depth, where your own product is one honest entry among fairly assessed alternatives, not a foregone conclusion - Fresh, machine-readable dates and real revisions, the one signal that held up under normalization in the 60,000-query data - Editorial independence a reader can feel, since the pages losing visibility are the ones that read as sales collateral wearing a list Are listicles dead for SEO? No. The unearned, self-ranking listicle is dying, and the useful one is being cited more than ever. The split runs straight through intent. Publish a list because it genuinely helps someone choose, and both Google and the AI engines still reward it. Publish one to rank yourself, and you are increasingly writing your competitor’s recommendation for them. That distinction is exactly the kind of quality rule worth encoding once and enforcing on every page, rather than trusting each writer to remember it, which is the same argument we made in AI search visibility beyond organic rank. If you are weighing whether to produce these pages at scale, our note on running a programmatic SEO pilot before scaling is the safer path than committing a template to thousands of URLs on a tactic this much in flux. Tags [#listicle-seo](/tags/listicle-seo)[#ai-search](/tags/ai-search)[#chatgpt-citations](/tags/chatgpt-citations)[#content-strategy](/tags/content-strategy) Share Written by [Carlos Arias](/authors/carlos-arias) Builder of AstroAgent, an AI-run website platform. On this page - The short answer: a trim at the top, not a purge - What the 60,000-query study actually measured - Where roundup pages still win: AI citations - Where the self-serving listicle tactic is collapsing - The verdict: build for the reader, not the scheme ## Continue reading [Reviews](/categories/reviews) · September 5, 2026 [### Listicle Ranking Signals: Freshness Beats Item Count](/blog/reviews/listicle-ranking-signals-research) Listicle ranking signals from a 60,000-query study: a stale date carried 56% lower odds of a top-3 spot, and which correlations stay unproven. Carlos Arias · 6 min [Reviews](/categories/reviews) · August 22, 2026 [### Answer Engine Optimization Jobs: The AI-Search Skills That Pay](/blog/reviews/answer-engine-optimization-role-signal) Answer engine optimization jobs now post at director level and six figures. The listings map exactly which AI-search skills carry market value. Carlos Arias · 5 min [Reviews](/categories/reviews) · August 18, 2026 [### Demand Generation AI Search Measurement: Six Frameworks](/blog/reviews/measuring-demand-generation-ai-search-era) Demand generation AI search measurement converges on brand mentions, entity coverage, and share of answer: signals developers can instrument. Carlos Arias · 5 min ## Stay in the loop. One email when it’s worth it — new posts and updates, no spam. Thanks — check your inbox to confirm. Free. Unsubscribe in one click.