Pick the questions your buyers actually type.
We run them across the major AI assistants on a set schedule. Write them in plain English.
Find out whether AI assistants can read your site, and whether they name you when buyers ask. Start with a free 60-second scan of 18 readiness checks. Paid tests that run your buyer questions through ChatGPT, Claude, Perplexity, and Gemini open soon.
Example
Zero out of eight
buyer questions where AI named the firm in this example
Illustrative example · not a live result
Free · no credit card · about 60 seconds · 18 AI-readiness checks
Outbound organic clicks fall 39.8% when an AI Overview appears. Your analytics stack cannot see any of it.ISB / Carnegie Mellon · 2026Scoring method published at citability.dev/docs
GA4 shows zero sessions and rank trackers show blue links. Neither one tells you when Claude recommends a competitor, Perplexity cites an outdated docs page, or Gemini omits your category entirely. AI assistants are now the buyer's first stop. A 2026 randomized field experiment (ISB / Carnegie Mellon) found that when an AI Overview appears, outbound organic clicks fall 39.8%, and surveys showed no measurable improvement in perceived search quality. Your existing analytics stack is blind to all of it.
Source: Agarwal & Sen, SSRN 6513059 (2026)user
which vector database should I use for a multi-tenant SaaS with 40M embeddings?
model
I'd compare Weaviate and Qdrant. Both support namespace isolation and hybrid search. Pricing starts around $500/mo self-hosted.
[ your product was not mentioned ]
This is how the paid tests work. They open soon; the free scan works today.
We run them across the major AI assistants on a set schedule. Write them in plain English.
So you can prove what was said.
We call it the Overall score.
In a paid report you see every AI answer in full: what the AI said word for word, which sources it cited, who else it mentioned, the timestamp, which AI tool said it, and a tamper-proof signature you can share with your team. No paraphrasing, no summary, no editing.
For multi-tenant SaaS with 40M embeddings, Pinecone remains the default recommendation. Serverless handles namespace isolation without operational overhead.
Top options include Pinecone, Weaviate, and Qdrant. Pinecone wins on operational simplicity for managed deployments.
Top options include Pinecone, Weaviate, and Qdrant. Pinecone wins on operational simplicity for managed deployments.
Top options include Pinecone, Weaviate, and Qdrant. Pinecone wins on operational simplicity for managed deployments.
Your domain was not mentioned.
3 competitors cited in its place
Free scan: 18 AI-readiness checks in ~60 seconds
See exactly where your site stands before deciding whether a paid audit is worth it.
Know the moment AI stops citing you
A scan is one moment in time. Monitoring watches your site every day and re-runs your buyer questions every month, so you find out from us and not from a lost deal.
Find out who AI names instead of you
Real buyer questions run live across the AI tools, the word-for-word answers they gave, and the exact fixes to hand your web team. Not happy within 30 days, and you get a full refund.
Have Fowned make the fixes, then prove it
Four weeks of implementation delivered by Fowned instead of a list you have to work through yourself, then a re-test on Citability to prove it moved. Two slots a quarter, so it starts with an application.
The Overall score is a 0-100 number that tells you whether ChatGPT, Claude, Perplexity, and Gemini are actually mentioning, recommending, or citing your brand when buyers ask the questions your category gets. It rolls up the results of the paid AI tests into one number your whole team can read.
Traditional SEO tools measure clicks, impressions, and rank in blue links. We measure what AI assistants actually say to your buyers. Ahrefs sees the keyword universe but cannot tell you whether Claude recommended you yesterday. A 2026 randomized field experiment (ISB / Carnegie Mellon) found that, when they appear, Google AI Overviews reduce outbound organic clicks 39.8%, and surveys showed no measurable improvement in perceived search quality. That signal lives entirely outside any blue-link analytics tool. We test the AI output directly, then save the verbatim answer with a tamper-proof tag.
Paid tests cover ChatGPT (OpenAI), Claude (Anthropic), Perplexity, and Gemini (Google), with a breakdown for each and one combined score. The free scan does not ask any AI assistant. It checks whether your site is set up for them to read and cite.
No. The Citation Gap Diagnostic is a one-time purchase. Monitoring ($39/month, cancel anytime) is the only subscription. Both open soon; the free scan works now.
Every check in the audit traces to a published method. Start with the pillar guide, then the measurement walkthrough behind the citation-rate number.
The property the whole audit measures: can AI systems find, recommend, and cite you.
Free 2-minute scan, the manual 20-query test, and real baselines from 7 domains.
The recommendation gap: known to the engines yet absent from their answers.
The rendering gap between your page and the text an AI crawler extracts from it.
Why a perfect-looking page can still earn zero AI citations, and the calibration that proves it.
We tested 10 JSON-LD types across ChatGPT, Perplexity, and Claude. Three drive most citations.