AIs combine methods such as query fan-out with classic search results

Explore Telemarketing Data Forum to share verified leads, sales strategies, and campaign tips. Connect with experts and boost your telemarketing success.
Post Reply
Sumona1030
Posts: 106
Joined: Tue Sep 23, 2025 3:25 pm

AIs combine methods such as query fan-out with classic search results

Post by Sumona1030 »

A recent analysis by Ahrefs provides interesting insights into the interaction between AI assistants and traditional search engines. According to the study, only around twelve percent of the links cited by ChatGPT, Gemini, or Copilot come from Google's top 10 results. On Bing, the rate is around ten percent. Perplexity stands out significantly with almost 29 percent – ​​thanks to its own index and crawler. Particularly striking: While Google AI Overviews rely on top 10 results 76 percent of the time, AI assistants often select sources independently of traditional rankings. This is due to methods such as query fan-out or reciprocal rank fusion, which spread content more widely.

At the same time
The Information reports that OpenAI allegedly luxembourg cell phone database accessed Google search results anonymously via the SerpAPI. Tests suggest that ChatGPT did indeed incorporate content from Google's results into answers – even without an official partnership. However, this finding doesn't necessarily contradict the study: ChatGPT can, on the one hand, use Google's data as a basis, but at the same time, it applies its own aggregation methods, which lead to a source selection that differs from that of traditional SERPs.

For SEOs, this paints a clear picture:

Top rankings in Google remain essential, especially for visibility in AI Overviews. At the same time, content beyond the first few results pages can also become relevant, as AI systems cite more broadly. Those who provide precise, structured content and work with clean technology increase their chances of being present in both traditional SERPs and AI responses.

Image


JavaScript rendering complicates visibility in AI search engines
Websites that rely heavily on client-side JavaScript rendering run the risk of being barely visible or not visible at all in AI-based search systems such as ChatGPT, Perplexity, and Claude. These platforms currently do not render content loaded exclusively via JavaScript—with potentially drastic consequences for the discoverability and display of websites in AI search.

An analysis by SEO Glenn Gabe clearly demonstrates this using a case study, confirming the findings of a 2024 Vercel study :
Content was not recognized by AI systems, favicons were missing, and the pages were either not cited at all or only in subordinate sections. Particularly problematic: Even targeted prompts and URL queries led to errors or empty results.

Website operators are therefore well advised to check their pages for JavaScript dependencies. Anyone who disables JavaScript in their browser and then notices that page content is missing or remains completely empty should consider a redesign. Server-side rendering (SSR) or reduced use of JavaScript for primary content are possible solutions. This lays the foundation for better visibility in the rapidly growing AI search market – and lays an important foundation for future GEO and LLMO strategies.

Linking Instagram Reels and translating them with AI – exciting innovations for creators
Instagram is introducing two new features designed to help content creators reach a wider audience and create more targeted content.

Reels can now be linked to each other. This allows content to be structured as a series and consumed seamlessly – including increased watch time and improved storytelling. Previously published Reels can be added to later. This improves user navigation, as viewers can be redirected directly to the next part, increasing engagement. A potential disadvantage, however, is that fewer users visit the profile – a previously common way to gain new followers.

Additionally, Instagram now enables AI-powered translations of Reels worldwide
Currently between English and Spanish. The creator's voice is simulated and lip-synced, creating a natural and authentic effect. This can significantly increase international reach. However, there are also concerns about data protection, potential misunderstandings due to incorrect translations, and the lack of labeling of such content as AI-generated.
Post Reply