Skip to content
Vouch
Join the waitlist
Menu

All posts

AI Search Visibility: What It Is and How to Measure It

By Vouch4 min read

Vouch blog cover with the title "AI Search Visibility: What It Is and How to Measure It" beside a line drawing of a gauge with its top point marked in red.
How the Vouch Rate is madeEach conversation scores from 0 to 1: half for how many of the six stages the brand survives, half for being recommended at the end, and sentiment moves the score by up to a tenth. The Vouch Rate is the week’s average times 100.Stages survivedHalfDepth of mention: 0 to 6 stagesChosen at the endHalfThe final answer recommends youSentimentUp to a tenthWarmer or cooler, up or downVouch Rate0 to 100weekly average
The Vouch Rate: depth of mention (stages survived), being chosen at the end, and a small sentiment adjustment, averaged over the week’s conversations into one number from 0 to 100.

AI search visibility is a simple idea with a messy reality: when buyers ask AI assistants about your category, does your brand show up, how is it described, and is it recommended at the end?

It is messy because there is no single results page to rank on. ChatGPT, Claude, Gemini, Perplexity, Google AI Overviews, AI Mode and Copilot each retrieve from different sources, cite different sites and recommend different brands. The same assistant rarely gives the same answer twice. Measuring visibility means measuring all of that, repeatedly.

The metrics that matter

Mention rate. In what share of answers is your brand named at all? Run your buyer prompts across assistants and count. This is the AI equivalent of "are you on page one."

Citation rate. How often does the answer link to your site as a source? Citations drive the growing stream of AI referral traffic, and they signal that the assistant treats your content as credible raw material.

Position and prominence. Named first in a shortlist or named fifth? Assistants tend to discuss the first options in more depth, and buyers anchor on them.

Sentiment and framing. Being named as "a budget option with reliability complaints" is visibility you do not want. The words around the mention shape the buyer's next question.

Recommendation rate. When the conversation reaches "which one should I buy," how often does the assistant choose you? This is the metric closest to revenue, and it can diverge sharply from mention rate.

Share of voice. Your mentions divided by all brand mentions in the category's answers. Ahrefs and Semrush have both shipped versions of this for AI answers, and it is the cleanest way to compare yourself with rivals.

Survival across the conversation. The newest and least measured: at which follow-up question does your brand disappear? Discovery, comparison, objections, trust, price, decision. A brand that survives all six stages is in a different league from one that wins the first answer and vanishes.

Why a single manual check misleads

Ask ChatGPT once whether it recommends you and you learn almost nothing. Assistants sample different sources on different runs, personalize by user and country, and shift tone run to run. Our own rule of thumb: treat any change as noise until it holds across several conversations and a few weeks.

That is why serious measurement runs each prompt multiple times per week, in each assistant, in each country you sell in, and tracks the trend rather than the snapshot.

How to build your baseline this week

  1. List 15 to 50 buyer prompts. The real questions buyers ask: "best X under Y," "X vs Y," "is X worth it," "X alternatives." Autocomplete in Google and the questions your sales team hears are free, reliable sources.
  2. Run them across assistants. At minimum ChatGPT, Perplexity and Gemini, plus Google AI Overviews where you rank. Note that assistants disagree: research consistently finds they cite different sources and name different brands for the same query.
  3. Ask the follow-ups. Do not stop at the first answer. Ask to compare the options, ask about downsides, ask which one to buy. Record where you appear and where you drop out.
  4. Log it weekly. A spreadsheet is enough to start: prompt, assistant, mentioned, cited, recommended, notes on framing.
  5. Connect it to traffic. In GA4, segment referral traffic from chatgpt.com, perplexity.ai, copilot.microsoft.com and gemini.google.com. AI referrals are still a small share of traffic for most sites, but they convert well and they are growing fast.

What good looks like

A healthy baseline is not "we are mentioned everywhere." It is knowing, with numbers: which assistants mention you, on which prompts, how they frame you, which rival gets recommended instead, and at which question you lose the conversation. From there, every content and PR decision gets sharper.

This baseline is exactly what Vouch automates. It runs your prompts daily across seven assistants, follows the six-question buying conversation on Claude, Gemini, ChatGPT and Perplexity, and shows your mention, survival and recommendation rates as a single Vouch Rate per week.

Get your baseline with Vouch: join the waitlist, and every invitation starts with a 14-day free trial.

Read next