A study across ChatGPT, Gemini, Perplexity, Claude, and Google AI Overviews found New Balance cited 1%→90% and lululemon 90%→0% by switching the category term from athleisure to athletic footwear; AI citation depends entirely on whether third-party content uses the same category language as the searcher.
Why it matters: GEO audits must include category-language mapping — if third-party coverage doesn’t match the framing customers actually use, no amount of entity optimization recovers the gap.
The method: compare JSON-LD entity declarations against vectorized content to surface coverage gaps, prioritize by business value, then monitor brand appearance across AI platforms with prompt-based tracking.
Why it matters: Brands with incomplete entity declarations have AI citation blind spots that keyword optimization can’t fix — entity gap audits are a distinct deliverable category worth pricing into engagements.
Mueller and Splitt clarified that 404s, unindexed pages, and redirect entries are often expected states; the right diagnostic is whether the indexing pattern matches intent, not whether every flagged item is cleared.
Why it matters: Practitioners can use this to reset client expectations around Search Console error counts — shifting from fix-it-list mode to trend review redirects effort toward content quality.
Reddit appears in 3% of AI citations (80%+ on comparative queries), search interest in entity SEO has grown 1,000%, and less than 8% of marketers still treat backlinks as a numbers game — the authority measurement layer has expanded well beyond hyperlinks.
Why it matters: Link building strategy now has to account for community presence, unlinked brand mentions, and entity relationships — earned media and brand hygiene are citation infrastructure, not optional add-ons.