Ahrefs has more than doubled the prompt dataset behind Brand Radar, expanding it from roughly 14 million to 30 million prompts. The source remains Google's People Also Ask data, but Ahrefs has lowered the search-volume threshold used to select questions.
The practical result is substantially deeper coverage of lower-volume and long-tail questions — precisely the kinds of natural-language queries that can matter in generative search even when their traditional keyword volume is small.
From roughly 14 million to 30 million prompts
Ahrefs says Brand Radar's prompt dataset grew from around 14 million to 30 million prompts. Rather than changing the underlying source, the expansion comes from moving further down the demand curve and admitting questions with lower search volume.
People Also Ask remains the source
The prompts continue to originate from Google's People Also Ask questions. Ahrefs uses these search-backed questions as a proxy for the kinds of natural-language requests people may also make to AI assistants.
This is an important methodological distinction: Brand Radar is not simply generating millions of hypothetical prompts. The dataset is anchored to questions associated with observable search demand.
Lower-volume questions bring more niche topics into the dataset
Lowering the volume threshold expands coverage beyond higher-demand questions. That can expose more specific product comparisons, niche informational needs and long-tail questions that previously fell outside the tracked prompt universe.
For GEO analysis, this matters because conversational AI usage is naturally compatible with detailed questions. A query may be strategically important even when its equivalent Google keyword has relatively little monthly search volume.
Country distribution remains broadly similar
Ahrefs says the country distribution of the expanded dataset remains roughly unchanged. The increase therefore represents deeper prompt coverage rather than a major geographic rebalancing of the database.
About 5% of prompts refresh each month
The company also says approximately 5% of the prompt pool is refreshed monthly.
That means the public dataset combines a large persistent prompt universe with partial monthly renewal rather than rebuilding the complete set every month.
The NetContentSEO view
This expansion is significant because AI visibility measurement becomes more useful as it moves beyond head terms. Long-tail questions are where niche publishers and specialist content can have a realistic opportunity to enter AI answers.
A larger prompt universe can therefore reveal visibility patterns that smaller, high-volume-only datasets miss. It also reinforces an emerging difference between classic rank tracking and GEO: low-volume questions can still matter when an AI system combines, reformulates or retrieves information across a broader conversational journey.
One caveat is important. Ahrefs' current Brand Radar documentation now describes a much larger overall AI visibility index than the 30-million-prompt expansion announced in this July/August update. The 30 million figure should therefore be read as the scale of this specific expansion milestone, not necessarily the present total size of every Brand Radar index.