3 builders who never spoke to each other

Earn the mention on a page you do not own

Language models read text rather than link graphs, and they discount a best-tools list published on the vendor's own domain because self-recommendation is easy to spot. So chase unlinked brand mentions on third-party sites: find the listicles already ranking in your category, contact the owner and ask to be named in the text. Years of answering questions in public now feed the same machine.

The method

1. 01

   Chase unlinked brand mentions on third-party sites alongside backlinks, because language models read text rather than link graphs and learn which names to cite.

2. 02

   Treat ranking and being cited as two separate disciplines: links move domain authority, unlinked mentions move your odds of being named by an AI answer.

3. 03

   A best-tools listicle hosted on your own domain gets heavily discounted, because the model can tell self-recommendation from a third party recommending you.

4. 04

   Getting cited in AI search takes coverage on other people's sites, because your own pages are only a fraction of what the models quote about you.

5. 05

   Answering questions on Reddit over a period of years now pays off indirectly, because AI assistants draw on those threads when recommending tools.

What it returned

• 75,000

  An Ahrefs study covering 75,000 brands put the tie between brand mentions and being surfaced by AI at roughly three times that of links.

• 85%

  Figure given: third-party sites are the origin of 85% of the brand mentions that surface in AI search.

• Years of Reddit Q&As now shape AI recommendations for Transistor.

Sources

• Hand-picked SEO conferences to attend in 2026 - Detailed.comMateusz Makosiewicz · detailed.com · Ahrefs Evolve: The future of discoverability · 2026-03-15
• Where are these signups coming from?Justin Jackson · justinjackson.ca · Where are these signups coming from? · undated
• SEO is NOT dead.Tibo Louis-Lucas · tmaker.io · What's actually happening · 2026-07-02