• by Laurie Sullivan , Staff Writer @lauriesullivan, April 15, 2025

Citations lacking in AI search results have become a major concern for publishers and brands looking for clicks to their web or ecommerce sites. 

Google executives have long said that links to pages will be added to AI Overviews when its technology determines it might be useful.

BrightEdge noted in March that specialized content within a website may be key to winning citations in Google’s AI Overviews. But a more recent study from Xfunnel of 768,000 citations shows how content strategies appear in engines and how they differ by category and region, as well as between B2B and B2C content.

Product-related content in AI search engine results -- including best-of -- topped the list in AI citations, at between 46% and 70% of all sources referenced.

The in-depth analysis published by Xfunnel shows how different content formats and structures influence artificial intelligence (AI) search engine citations and recommendations. The company helps businesses optimize their presence across AI search engines like ChatGPT, Perplexity, and Claude.

The research analyzed the most often cited content when answering user questions. The company conducted the study during a span of 12 weeks, analyzing content from ChatGPT, Google AI Overviews, and Perplexity.

Content that struggled to gain citations as often as products included news and research, affiliate, user reviews, blog and PR materials.

News and research articles each received between 5% and 16% of citations, affiliate content remained below 10%, user reviews like Q&A sites came in between 3% and 10%, blog content received between 3% and PR material garnered less than 2%.

Xfunnel’s analyses also looked separately at B2B and B2C queries.

B2B queries lead to fewer, more authoritative sources, according to the data.

Nearly 56% of citations for B2B queries in search were for product pages such as company or vendor sites. Affiliate received 13% and user reviews received 11%. News followed with about 9% and research with about 6%.

The analysis for B2B suggests that it highlights a strong reliance on first-party resources in business contexts, particularly for technical or enterprise-level questions.

B2C queries produced a broader mix, with more voices from affiliates, review sites, and general media.

Product content for B2C queries in search came in at about 35%, doing a bit better than B2B. Affiliate received 18% and user reviews received 15%. News followed with 15%.

Citations differ between regions, but each “adapted to the availability and perceived reliability of local content,” which made the findings interesting.

In North America about 55% were product citations, while news came in at about 10%, and research at about 10%. Product citations in Europe came in at about 50%, while product citations in Asia Pacific came in at about 45.9%, and Latin America had about 62.6%.