Self-Promotional Listicles: Is Your “Best [Category]” Page Recommending Your Competitors?

If your marketing team has built a “best [your category]” page with your own product sitting at number one, you’ve published what the industry now calls a self-promotional listicle. For about a year it was standard GEO advice. New data suggests it may be handing your sale to someone else. In one analysis of 100 B2B queries, Google’s AI Overviews cited the brand’s own list most of the time, then recommended other companies in the answer text.

That’s a strange outcome. You did the work, you got the citation, and a competitor got the pick. Below we walk through what the research found, why it happens, what it might mean for your organic traffic and your legal exposure, when a comparison page is still worth building, and what to do instead. This sits inside the broader content marketing work we do, and it picks up a thread from our look at brand mentions versus citations.

What are self-promotional listicles, and why did so many brands publish them?

A self-promotional listicle is a “best X” or “top 10 Y” article written by a company that ranks its own product in the list, usually near the top. They spread because list-style pages get cited a lot by AI systems, and plenty of agencies told clients to make them.

The citation numbers explain the appeal. Evertune Research looked at roughly 25,000 URLs pulled from the most-cited pages across six AI systems (ChatGPT, Copilot, Gemini, Google AI Mode, Google AI Overviews and Perplexity), using March and April data. Half of those URLs were listicles, and listicles made up 63 percent of nearly 400 million citations. Ranked lists such as “Top 5 CRM Tools” were 71 to 86 percent of the listicles. One caution: Evertune sponsored that piece on Search Engine Land, and the site says it doesn’t confirm or dispute the conclusions, so read it as a vendor study.

The same wave is visible on the supply side. Lily Ray found 184 self-promoting listicle pages across 146 brands in her AI Overview data, with a surge through 2025. Her read is that this is a sample of something much bigger.

Why does Google cite my “best” page but recommend my competitors?

Because a citation and a recommendation aren’t the same event. The AI Overview can pull facts from your page, show it in the source list, and still name other companies as the answer. In Ray’s data that happened in 224 of 323 cases, or 69 percent.

Here’s how she set it up. She took 100 B2B “best [category] software” queries, one per category, and checked them on April 15, May 15 and June 8, 2026, using Ahrefs Brand Radar. Eighty of the 100 prompts returned an AI Overview. “Cited” meant the brand’s list showed up in the sources. “Recommended” meant the brand was named as a pick in the answer itself.

The brands that did get recommended looked alike. They were already category leaders, with more referring domains, more mentions across the web and stronger link profiles. In project management software, for example, Asana, Monday, ClickUp and Notion were recommended while smaller tools were cited without the nod. In CRM for small business, only Monday was recommended, a self-promoter with a Domain Rating in the 90s and about 75,000 referring domains. Ray is upfront that the cause is inferred, not proven, and a few lesser-known brands still got picked now and then. Still, the pattern is plain enough. If your page lists the market leaders next to you, the answer engine may simply believe the leaders.

Her practical point: stop counting citations as the win. Track whether you’re named. If you haven’t set that up yet, our piece on mentions versus citations shows how to separate the two scoreboards.

Are self-ranked listicles hurting my organic traffic too?

They might be. The honest answer is that nobody has proven it. What exists is a pattern: sites with lots of these pages lost visibility starting around January 19 to 21, 2026, and the declines continued through Google’s May 2026 core update.

In her February analysis, Ray reviewed seven anonymized sites and found Sistrix visibility down between 29 and 49 percent on six of them. One site had 191 listicles inside about 30,000 blog articles. Another had 228 out of roughly 2,780. A third had 267 out of 2,790 pages, 76 of which carried “2026” in the title. Later, in her June write-up, she described the drops reaching whole domains on more than 40 sites, not just the list pages.

Now the caveats, because they matter. Some badly hit sites in her set used few listicles and relied on other repeatable page templates instead. The affected sites also shared other traits, like fast content scaling, suspected AI-generated text and dates that were refreshed without real edits. Google hasn’t announced a crackdown. A Google spokesperson did tell The Verge, as relayed by Search Engine Roundtable on April 14, 2026, that the company has “robust protections against common forms of manipulation” in Search and Gemini and is aware of low-quality listicle content. So treat this as a warning sign, not a verdict. If you’re running dozens of these, a content review is cheaper than finding out the hard way. We cover that process in our guide to auditing and pruning content for AI search.

Could a self-ranked “best” list cause legal trouble?

Possibly, if the page reads as an independent review when it isn’t. The FTC’s Consumer Review Rule bars a business from falsely claiming that a site or entity it controls offers independent reviews or opinions. Whether your specific page crosses that line is a question for a lawyer, and we aren’t one.

The FTC announced the final rule on August 14, 2024. Per the agency, a company can’t misrepresent a website it controls as an independent source of reviews for a category that includes its own products. A law firm client alert from Venable, dated December 29, 2025, describes the same prohibited conduct: running a seemingly independent review site without clearly disclosing the company’s ownership or control. Search Engine Land has reported that the risk rises when a page implies objectivity while promoting the publisher’s own product, uses invented scoring, or lists competitors that were never tested. The FTC’s inflation-adjusted maximum civil penalty for this kind of violation was $53,088 per violation as of its January 2025 adjustment, and that figure is updated periodically. Search Engine Land’s coverage suggests each page could be counted separately, which is the part that should make a team with 200 listicles stop and think.

None of that means every comparison page is illegal. It means clarity is the dividing line. A page that says “we built this, here are our criteria, here’s where a competitor beats us” is a very different thing from a page dressed up as a neutral roundup.

When is a comparison page still worth publishing?

When it serves a buyer who is actually choosing between named options, tells the truth about the trade-offs, and doesn’t pretend to be independent. That’s our view, based on the research above rather than a rule from Google.

A quick test we’d run before building one:

1. Is your brand already a recognized name in the category, with third-party coverage to back it up? If not, a self-ranked list is likeliest to promote someone else.
2. Can you state your criteria and show the work? Real testing and stated methodology beat a vibes-based ranking.
3. Will you say where a competitor is the better fit? If every row favors you, readers and machines both notice.
4. Is the page labeled as written by your company, in plain words near the top?
5. Would you be comfortable showing it to your legal team? If the answer is no, don’t publish it.

If you pass most of those, a clearly labeled “X vs Y” or “how to choose” page is a reasonable asset. If you fail the first one, put the effort somewhere else.

What should I build instead to get recommended?

Get named on pages you don’t control, and publish evidence that others can’t copy. Ray’s data shows Forbes, Reddit and YouTube among the most-cited domains for “best” queries. That tells you where the recommendation signal lives, and it isn’t on your blog.

Here’s the order we’d work in:

1. Earn the third-party list. Pitch the publications and review sites already ranking for your category. Give them a reason to include you: a product demo, a customer result, a fact they can verify.
2. Publish original data. Surveys and benchmarks give journalists and AI systems something to cite you for. See our write-up on original research and AI citations.
3. Fix what your site claims. Check every stat and superlative on your pages before a model repeats it. Our fact-checking guide for AI-assisted content walks through it.
4. Measure the pick, not the cite. Track whether the AI answer names you for your target queries, and log it monthly.
5. Retire the weakest self-ranked pages. Don’t mass-delete overnight. Review them, merge what’s useful into honest comparison content, and redirect the rest.

It’s slower than publishing a listicle. It’s also the only version where the recommendation can’t be taken away by a policy change.

Sources

• Lily Ray, “Why Calling Yourself the Best Could Be Helping Your Competitors Win in AI Search,” Substack, June 17, 2026 (100 queries, Ahrefs Brand Radar, checkpoints April 15, May 15 and June 8, 2026).
• Lily Ray, “Is Google Finally Cracking Down On Self-Promotional Listicles?” Search Engine Journal, February 5, 2026 (Sistrix U.S. Visibility Index; seven anonymized sites).
• Evertune Research, “AI search loves listicles: what 25,000 URLs reveal about citations,” sponsored content on Search Engine Land, 2026 (page carries no stated publication date; data covers March and April). Sponsor-authored.
• Mordy Oberstein, Search Engine Roundtable, April 14, 2026, relaying a Google spokesperson statement to The Verge.
• Federal Trade Commission, press release announcing the final rule banning fake reviews and testimonials, August 14, 2024 (16 CFR Part 465).
• Venable LLP, “FTC Signals Heightened Enforcement of New Consumer Review Rule,” December 29, 2025; FTC, inflation-adjusted civil penalty amounts, published February 11, 2025 ($53,088 per violation). Not legal advice.
• The section questions above are common query patterns for this topic; we did not verify search volume for each one.