If you've been searching for your brand in ChatGPT or Perplexity and coming up empty — even on topics you know you rank well for in Google — you're not alone. Strong SEO doesn't translate automatically into AI citations. They're driven by different signals. Here are the four most common root causes, and exactly what to do about each.
Root Cause 1: AI Engines Don't Know Your Brand Exists as an Entity
AI models build knowledge of the world through training data. If your brand doesn't appear consistently across authoritative sources — Wikipedia, Crunchbase, Wikidata, LinkedIn, industry databases, press coverage — the AI may have no confident 'entity' for your brand in its internal knowledge. It's not that the AI disagrees with you. It's that it doesn't have enough signal to confidently reference you at all.
The fix: establish entity presence systematically. Secure a Wikidata entry (straightforward for any real company). Ensure your brand appears in your industry's primary databases (G2, Capterra, ProductHunt, Crunchbase). Get at minimum a Wikipedia mention in a relevant category article, if not your own entry. Link all of these with the sameAs property in your Organization schema.
Root Cause 2: Your Content Doesn't Answer the Questions Being Asked
Great SEO content is often long, authoritative, and comprehensive. But AI engines specifically retrieve content that directly answers a question — ideally in the first sentence or two. If your content buries its point, takes 500 words to get to the answer, or is structured as thought leadership rather than direct response, it's a poor retrieval candidate for AI engines even if it performs well in search.
The fix: audit your top pages for direct answer structure. For every page you want to rank for, identify the primary question it should answer, then rewrite the opening to answer it directly. Add a FAQ section that covers the 5–8 questions customers actually ask AI engines about your topic. Layer FAQPage schema on top of it.
Root Cause 3: No Structured Data (or Broken Schema)
This is the most fixable cause — and the most commonly overlooked. Sites with zero JSON-LD schema ask AI engines to infer what they're about from unstructured text. That's a much higher bar. Sites with broken or inconsistent schema (where the schema contradicts the page content, or uses deprecated properties) are often penalised rather than helped.
The fix: run a full schema audit. Check every key page for: the presence of JSON-LD, correct typing, required properties, and consistency with on-page content. Prioritise Organization on homepage, Article on blog posts, FAQPage on any Q&A content, and Product/Service on commercial pages.
Root Cause 4: You're Only Tracking the Wrong Queries
Many brands discover AI citation problems late because they're only monitoring brand-name queries ('Is [Brand] good?') rather than category queries ('What is the best [category] tool?'). AI citation matters most on category, comparison, and informational queries — the ones where a buyer is still forming an opinion and your brand needs to be in the answer to enter consideration.
The fix: build a query set that mirrors real customer intent at each stage of the buying journey. Include: category queries ('best [category] software'), comparison queries ('[Brand] vs [Competitor]'), use-case queries ('how to do [thing your product does]'), and problem queries ('how to solve [pain point]'). Monitor these queries regularly across ChatGPT, Perplexity, Gemini, and Claude.
Most brands have one primary cause and one secondary cause. Fix the primary one first — usually entity establishment or structured data — and you'll often see meaningful movement within 4–6 weeks.
How Long Does It Take to See Results?
AI citation improvement timelines vary. Schema changes that are picked up by search crawlers typically propagate within 2–4 weeks. Content changes take slightly longer — 4–8 weeks — as AI retrieval systems re-index and re-weight your pages. Entity establishment (Wikipedia, Wikidata, press coverage) takes the longest but has the most durable impact.
The key is to measure systematically so you know which changes are working. That's what Citingly is built for — tracking your citation rate across engines over time, correlating it with the changes you make, and surfacing the next highest-impact action.