Self-promo listicles: GEO upside, backfire risk
Self-promotional "best of" listicles are one of the most popular GEO tactics: brands publish comparison articles, rank themselves first, and hope for citations in ChatGPT, Google AI Overviews, or Perplexity. Two recent studies paint a fragile picture. The strategy works in the short term, but over time it can cost visibility and hand an advantage to competitors.
The starting point is a large Ahrefs analysis by Glen Allsopp. He examined ChatGPT responses to 750 top-of-funnel prompts in three categories: software, products, and agency recommendations such as "the best web design agencies in London." The study covered 26,283 source URLs. The core finding: "best [category]" listicles accounted for 43.8 percent of all cited page types, far ahead of landing pages, product pages, or classic blog posts. AI systems rely heavily on structured comparison lists for recommendation queries.
Why listicles dominate AI answers
Listicles deliver exactly the format language models need for answers: numbered options, short descriptions, clear rankings. Allsopp also found that 79.1 percent of cited best-of lists were updated in 2025, a strong freshness signal. Brands placed higher in third-party lists appear more often in ChatGPT recommendations. Even when a company puts itself at the top of its own list, the page can still appear as a source; in the software category this happened in more than one third of cases, and even more often in the agency category.
Many SaaS and B2B brands scaled this insight programmatically: hundreds of "best X 2026" articles, often with superficial year updates instead of real retests. Major names like Shopify, HubSpot, and Slack used the format too. The short-term logic seemed sound until pushback from traditional search made the line between GEO success and SEO risk visible.
Lily Ray: citation is not recommendation
The second study by Lily Ray, vice president of SEO strategy at Amsive, focuses on Google AI Overviews. Using Ahrefs Brand Radar, she analyzed 100 B2B queries matching "best [category] software" at three checkpoints between April and June. In 80 prompts that triggered an AI Overview, self-promotional listicles appeared as sources 323 times. In 224 cases, Google cited a brand's own listicle page but did not recommend the brand itself, roughly 69 percent of the time.
In practice, a brand writes the comparison, gets used as evidence, but customers are sent to established competitors named in the same article. For LMS, CRM, or help desk software queries, the pattern repeated: the listicle page supplied structure while the recommendation went elsewhere. Visibility in the AI answer and conversion proximity are two different metrics that must be measured separately.
Typical backfire patterns
- Citation without brand recommendation in AI Overviews
- Organic visibility losses of 30 to 50 percent in blog and guide folders
- Losses concentrated in self-promotional best-of content
- Spillover to AI Mode, Gemini, and indirectly ChatGPT through index dependency
Google's pushback on manipulative lists
Ray documented steep losses at dozens of SaaS sites starting in mid-January 2026, partly accelerated by Google's May core update. Blog and resource sections heavy on self-promotional listicles were hit hardest, not always the full domain. Google's review guidance emphasizes first-hand experience, original data, and transparent methodology. Lists without independent testing, without bias disclosure, and built mainly to influence rankings conflict with those signals.
Common patterns among penalized sites: programmatic templates at scale, aggressive year refreshes without substantive change, and AI-generated copy without editorial depth. The effect cascades: brands that drop in Google often lose ground in AI Overviews and related surfaces because they lean heavily on the search index. For GEO teams, a ranking shortcut in classic search can drag down overall AI visibility.
| Study | Key figure | Implication |
|---|---|---|
| Glen Allsopp / Ahrefs | 43.8% listicle share across 750 prompts | Format is heavily cited by ChatGPT |
| Lily Ray / Amsive | 69% citation without brand recommendation | Own listicles often boost competitors |
| Ray visibility analysis | 30-50% loss in blog folders | Scaled self-promo lists turn risky |
What SEO and GEO teams should check now
Teams should measure not only whether a URL is cited but whether their brand is recommended. Brand Radar tools, AI Overview monitoring, and comparing citation versus recommendation rates are essential. Audit best-of pages for real testing methodology, disclosure of self-interest, and independent evidence. Third-party placements in credible industry lists often outperform self-ranked number-one spots.
- Add "how we tested" sections with traceable criteria
- Assess self-promo listicles by scale and risk, not only short-term citations
- Prioritize third-party mentions and genuine reviews as GEO levers
- Track blog subfolder visibility separately after algorithm volatility
For publishers with large content hubs, the math gets worse: when blog folders carry three quarters of organic visibility and the listicle strategy is concentrated there, a single algorithmic shift can threaten most of the traffic. Monitoring must run at subfolder level, not only domain-wide.
The data shows a clear tension: self-promotional content works until it backfires. Teams that treat GEO as a ranking shortcut risk becoming sources for AI systems while competitors capture the recommendation that actually matters.
