SyncAI.news, a Varaisys broadcasting
Why Your Brand’s Next Visibility Battle Is Being Decided In AI Answers, Not Search Rankings
OF

Ofer Familier, Forbes Councils Member

· 1 min read

World NewsForbes: Innovation

Why Your Brand’s Next Visibility Battle Is Being Decided In AI Answers, Not Search Rankings

Ofer Familier, Co-Founder and CEO of Dig.

Until recently, finding recommendations online meant working through a page of search results. Now, people ask an AI and get a clear list back. What counts as visibility has changed, from clicks, impressions and rank to whether a brand shows up in an AI answer and how.

The instinct is to treat this like SEO with a new name, and brands risk losing ground because of it. An AI-generated answer doesn’t simply grab the top two Google results and hand them back. That’s the gap that explains why a brand can rank well on Google and still end up missing, misdescribed or left out of the exact conversation that shapes how people see it.​

Why SEO’s Tool Kit Doesn’t Carry Over

Generative engine optimization (GEO) works entirely differently from SEO at the root. SEO rewarded climbing a ranked list through keywords and backlinks. GEO is the practice of shaping how a brand gets represented in an AI-generated answer.

The processes of SEO and GEO aren’t the same, and for GEO, ranking well based on keywords doesn’t reliably help the way it used to. Brands now need to identify where AI builds its answers from, and the sources are varied, including reference sites like Wikipedia, publications and journals, review and comparison platforms, owned websites and even social media, depending on the system. How AI systems draw on these platforms isn’t always clear.

ChatGPT, Gemini and Claude source differently. ChatGPT includes a citation in 96% of its answers but keeps the list short, averaging around five sources per response, while Gemini cites slightly less often, in 82% of responses, pulling from about eight sources when it does. Claude is the outlier, citing in only about half its responses, but going deeper each time, averaging around 13 sources. Where each system tends to look also differs, with ChatGPT using Wikipedia as its most common source, Claude using PubMed Central and Gemini favoring Reddit.

Original source

This story was published by Forbes: Innovation and written by Ofer Familier, Forbes Councils Member. SyncAI.news shows a preview; the complete article is on the publisher's site.

Read the full story on forbes.com

Similar News