BLOG

Notes from the answer layer.

How AI engines choose their sources, what moves the answers, and what we're learning from the responses we store.
Latest
Measurement7 minRankings flat but traffic falling? Here is whyImpressions hold while clicks fall because AI answers resolve the query in place. How to confirm search decoupling, and what to measure instead.Practice8 minHow to get cited by ChatGPTChatGPT cites what it retrieved, often not the pages ranking for your query. What earns a citation, and which formats get lifted whole.Data6 minWhich content formats do AI engines cite most?Across 286 competing citations, comparison pages won 46% while the brand publishing them held 6%. What that gap says about format versus angle.Practice7 minHow to appear in Google AI OverviewsAI Overviews cite a small set of sources that are frequently not the pages ranking beneath them. What decides inclusion, and why ranking first is no guarantee.Data5 minWhy AI answers disagree with each otherEight major models agreed on their top recommendation only 43.9% of the time. What that means for anyone checking their brand in a single assistant.Measurement6 minHow to measure AI share of voiceAI share of voice is decided by its denominator. Count only tracked competitors and your score rises whenever you remove one.Practice7 minWhy ChatGPT names your competitor, not youAbsence from an AI answer is usually a retrieval problem, not a content problem. The four causes, in the order worth checking them.Measurement5 minHow often should you track AI visibility?Daily tracking mostly measures variance. Why weekly is the honest default for most brands, and when daily genuinely earns its cost.Practice4 minDo you need an llms.txt file?No major engine guarantees support. Publish one anyway, for a reason that has nothing to do with rankings.Practice6 minShould you block AI crawlers?Blocking GPTBot protects your archive and removes you from retrieval. Which businesses should, which should not, and how to do it selectively.Measurement7 minHow to build an AI visibility prompt setThe prompt set decides everything downstream. How to choose questions buyers actually ask, and the mistake that quietly ruins the whole programme.Practice6 minWhat to do when ChatGPT describes your brand wrongAssistants state outdated prices, missing features and wrong compliance claims. You cannot edit the answer, so you change what it reads.Measurement5 minHow to report AI visibility to your CEOFour numbers, one chart, and the caveat that keeps you credible when the score moves for reasons nobody controls.
FREE VISIBILITY REPORT

The most useful thing we publish is about you.

Enter your work email and we'll run your first visibility scan: per-engine scores, top competitors, and the sources they're cited from.
7 DAYS FREE · FIRST SCAN TODAY
volexi
MONITOR · ACT · MEASURE
The AI search console for marketing teams and the agencies that serve them.
© 2026 FOXTED LIMITED T/A VOLEXIGDPR-READY · EU DATA RESIDENCY