15 Million Data Points Say AI Search Rewards Boring SEO. The Hacks Can Sit Down.
Ahrefs analyzed 15 million data points to determine what actually influences rankings in AI search. The findings confirm that slower, familiar strategies still win: proper ranking, earning genuine mentions, building topical content depth, and creating a brand people actually discuss. AI search is described as allergic to hacks. One specific myth busted: publishing 'best of' lists featuring your own brand does not reliably trick AI into recommending you. Across 750 ChatGPT prompts studied by Glen Allsopp, 'best X' blog posts made up 43.8% of all page types referenced.
This demonstrates the principle of signal durability. The mechanism is that AI search engines synthesize consensus from multiple sources, meaning manipulation of a single source has diminishing returns. What persists across algorithm changes is genuine brand recognition and topical authority. The lesson: if your strategy only works because a system is imperfect, you are one update away from irrelevance. Build things humans cite because they are useful, not because they game a ranking.
Ahrefs, which conducted the analysis of 15 million data points, and Glen Allsopp, who studied 750 ChatGPT prompts and found that 'best X' list posts accounted for 43.8% of all page types referenced by the AI.
- Open ChatGPT or Perplexity and ask it 'What are the best [your industry or niche] tools or services?' Read the full response carefully and note whether your brand appears anywhere.
- Search for your brand name plus the word 'review' or 'recommended' on Google. If nobody independent is mentioning you, that is your actual problem. AI search aggregates human consensus, so you need real mentions first.
- Pick one topic adjacent to your business where you have genuine expertise and write a thorough, useful resource on it. Post it on your site or a platform like Medium. The goal is to become the source that other humans cite, because that is what AI search ultimately surfaces.