How do I know if AI is recommending my brand?
The three outcomes you must track separately
Most tools and most agencies collapse these into a single 'visibility' number, which hides the information you actually need.
A citation means your domain appears as a linked source in the answer. This is the most valuable outcome because it produces referral traffic as well as credibility.
A mention means your brand is named but nothing links to you. This has real value — it builds familiarity and often precedes citation — but it generates no traffic, so counting it alongside citations inflates apparent performance.
A recommendation means you are named as a preferred option rather than merely referenced in passing. This is the outcome that actually drives revenue, and it is the hardest to earn. A brand can be cited frequently as a source while never being recommended as a choice.
Building the prompt set
The prompt set is the instrument, and its design determines whether your measurement means anything. Three rules matter.
Write prompts the way buyers actually write them. Not 'best CRM' but 'we're a 12-person agency outgrowing spreadsheets, what CRM should we look at'. Answer-engine queries are conversational and situational, and the answers they produce differ substantially from those returned by keyword-style prompts.
Cover the whole funnel. Category-definition prompts, comparison prompts naming your competitors, objection prompts about price and reliability, and direct purchase-intent prompts. Each behaves differently and each is worth knowing about.
Then freeze it. The single most common measurement error is rewriting prompts between cycles, which makes it impossible to distinguish your progress from model drift. Between 40% and 60% of AI citations change monthly on their own — you need a fixed instrument to see through that noise.
Why one engine is not enough
Profound found that only 11% of domains appear in both ChatGPT and Perplexity. The overlap is far smaller than most people assume, which means measuring a single engine tells you very little about your position in the others.
The engines also draw on different retrieval mixes, and those mixes move. Between April and July 2025, ChatGPT's alignment with Google's index rose from 12% to 33% while its alignment with Bing fell from 26% to 8%. Any strategy tuned to a single engine's retrieval behavior is standing on ground that shifts quarterly.
Attribution: separating AI traffic from organic
Most analytics setups fold AI referrals into direct or organic traffic, which means the channel appears not to exist. Filtering by referrer for chatgpt.com, perplexity.ai, gemini.google.com and claude.ai gives you an isolated view.
Do this before you start any AEO program, because the volume will look disappointing and the conversion rate will not. Seer Interactive measured 15.9% conversion on ChatGPT referrals against 1.76% for Google organic. Judged on sessions the channel looks marginal; judged on customers acquired it frequently is not.
Can I check AI brand visibility manually?
Yes, and it is worth doing once before buying tooling. Write twenty buyer questions, ask each of ChatGPT, Perplexity, Gemini and Claude, and record whether you are cited, mentioned, or recommended, and who is named instead. It is tedious but it produces a real baseline and tells you whether you have a problem worth paying to solve.
What is a good share of answer?
There is no universal benchmark because it depends entirely on category competitiveness and prompt breadth. The meaningful comparison is against your own baseline and against named competitors on the same fixed prompt set. Appearing in 30% of answers in a crowded category may be strong; appearing in 60% in a niche one may be weak.
How often should AI visibility be measured?
Monthly. Between 40% and 60% of AI citations change every month, so quarterly measurement misses most of the movement and cannot attribute changes to specific work. Weekly measurement mostly captures noise and adds cost without adding signal.
Tell us what you are building.
If the question is about discovery, reputation, a new site, or a market move, send the context. AIGNCI will tell you whether an Audit, a build, or a more focused engagement is the right starting point.