Who Gets Mentioned When the Machines Answer

Across the engines tracked by the panel, ChatGPT and Claude account for the largest share of brand mentions and show the highest thirty-day Mention Rates. The latest weekly readings put ChatGPT at 22.8% and Claude at 15.0%, after an eleven-week stretch in which Claude peaked at 32.9% and ChatGPT bottomed out at 6.5%.

Share reflects the distribution of mentions across engines, Mention Rate measures how frequently an engine mentions a tracked brand in its answers. ChatGPT holds the larger share of all mentions recorded; Claude is more likely to name a brand in any given answer. Gemini, DeepSeek, and Perplexity trail well behind on both counts.

Blogs continue to account for the largest share of citations in the panel. Social sources, however, have expanded considerably during the reporting period: Social's share rose from 2.7% to 14.9% — a five-and-a-half-fold increase. Over the same span, blogs slipped from 89.3% to 76.4%.

Blogs remain the dominant citation source, but their share has declined over the reporting period. Social sources and forums now account for a larger proportion of observed citations. For GEO teams, the findings support a broader approach to source development: maintain strong owned content while investigating relevant social platforms, forums, and publishing platforms.

The overwhelming majority of recorded citations came from sources classified as free to publish on. This is a description of the observed citation landscape, not a comparison of the success rates of free and paid placements. The size of the free web means that its share of citations cannot, by itself, establish which channel delivers the best results per outreach attempt.

RankCaster classifies citation opportunities by the potential action required to secure a brand's presence in a source. Editorial outreach is the largest category, accounting for 83.6% of classified action paths. Other categories include optimizing owned documents, self-publishing, editing wiki pages, submitting directory listings, academic submissions, review solicitation, and paid placement.

Perplexity has the highest Competitive Density in the panel, averaging nearly ten distinct tracked brands per answer. Gemini averages about five. Density describes how many brands appear in an answer; it does not indicate how prominently any individual brand is presented. Claude records the highest Share of Voice, followed by ChatGPT. Comparing the two metrics gives a more complete picture of brand presence: an engine may include many brands in its answers without giving the tracked brand a large share of total mentions.

Where the Opportunities Lie

Citation make-up · last 30 days · ≈138,000 citations classified by their relationship to the tracked brand.

Opportunity sources — pages where neither the tracked brand nor its competitors were detected — account for approximately half of classified citations. Competitor-associated pages make up another 38.9%, a second major area for source analysis. Own-domain citations represent 1.5% of the classified total; the figure describes the observed citation mix, not a ceiling on what an owned-content strategy can achieve.