Product

Visibility Sentiment Competitors Citations AI Traffic Opportunities Pricing

Engine trackers

ChatGPT Google AI Overviews Perplexity Claude Gemini Grok Meta Spark AI Mode

Resources

How it works Blog Learn Log in
AI Visibility

Track your brand's visibility in AI answers.

One score for how often AI mentions your brand: trended over time and measured against the competitors fighting you for the same answers.

Join the waitlist

7-day trial, no card

Latest scan 21 Aug 06:00
AI visibility
62.5% +3.3
Avg position
#2.1 +0.3
Named first
31 /75
Sentiment score
+34 +6
Visibility per engine
ChatGPT 68% +4
Perplexity 61% +1
AI Overviews 58% -2
Claude 54% +3
Gemini 49% 0
Meta Spark 44% +2
Grok 38% -1
AI Mode 35% +5
3 engines on every plan. 5 more as add-ons. 8 engines

Ask every engine

ChatGPT, AI Overviews and Perplexity on every plan. 5 more you can add.

Track the real questions

You pick the prompts, worded the way customers ask them. 15 to begin with, several hundred if you need them.

Read the tone

Every mention is classified positive, neutral or negative, with the exact sentences that earned it.

See it move daily

Answers change without warning, so every prompt runs once a day and every run is kept.

The basics

What is AI visibility?

A buyer used to type the question into Google, skim ten results, and click a few to make up their own mind. Now they ask an assistant and get one answer, and that answer names two or three brands. Whoever it leaves out never gets considered.

AI visibility is how often you're one of those names: the share of answers that mention you at all.

There's no page two. Either the answer names you or the buyer never hears of you, and which brands it names can change from one day to the next.

One question, one shortlist 21 Aug
Best trail running shoes for muddy ultras?
For muddy ultras, the standouts are Hoka Speedgoat for cushioning over distance, Salomon Speedcross for grip in deep mud, and inov-8 Mudclaw as the lightweight option.
You're named: 2nd of 3 3 brands named
Methodology

How the score is calculated.

A number is only worth trusting when you can see how it was reached. Three steps happen to every prompt, every day, and a handful of rules keep the result comparable week to week.

1 Ask

Once a day we send each prompt you track to every engine you have switched on, in your wording rather than ours.

best trail shoes for muddy ultras
8 engines Once a day
2 Read

We keep the answer verbatim, then read it for who got named and where you landed. The pages it leaned on get saved with it.

1 Salomon2 Hoka3 Merrell
3 Score

It is the share of today's answers that name you. Next to it, how much that shifted overnight.

+4.2% vs yesterday
The scoring rules 7 rules +
01
Once a day, per engine

We take one answer per prompt, per engine, per day. The score settles across engines and days rather than repeated runs of the same question.

02
Same wording every day

We send your prompt exactly as you wrote it. Rephrasing it to chase a better answer would leave you comparing two different questions.

03
Named means named

The answer has to say your name. Talking around the category you sell into earns you nothing.

04
Position counts

We read where you land among the brands named, first or fifth, as well as whether you turned up at all.

05
Scored per engine

You can score any engine on its own, or all of them at once. The blended number counts every answer once over.

06
Empty answers are results

Some questions come back with no brands in them. That answer still counts towards the day's total instead of dropping out of it.

07
Every swing is dated

We compare each score with the same prompt's answer from the scan before, so a swing carries the day it happened.

Which engine moved, and when.

Engines don't move together. Filter to one engine and every score on the page re-scopes to it, so a drop reads as somewhere to go and fix instead of disappearing into an average.

Salomon · Visibility by engine · 30 days
ChatGPT
72% +4%
AI Overviews
60% -2%
Perplexity
55% +6%
AI Mode
52% +2%
Claude
48% +1%
Grok
44% -1%
Gemini
41% -3%
Meta Spark
33% +5%

One score, everything behind it.

Visibility is the entry point, not the whole story. Your markets, the kind of demand you win, and the reason behind every move: each one click deep, none a dead end.

Every market, its own score

Winning in London says nothing about Berlin. Prompts are asked per market, in the local language, so each score reflects what customers there hear.

United Kingdom 62.5%
Germany 54%
Japan 38.5%

What kind of demand you win

Prompts are tagged by intent, so you can tell a branded search you were always going to win from an unbranded one you actually earned.

Unbranded 38%
Commercial 34%
Branded 28%

Every move is explained

Every scan is diffed against the last, per prompt, per engine, so a two-point drop reads as which answer changed.

Changelog · 21 Aug 2 moves

"…for muddy ultras?"

+2.5 pts

Perplexity

"best grip in wet rock?"

+1.1 pts

ChatGPT

Visibility questions, answered.

How the visibility score is built, and what moves it.

How is AI visibility calculated? +

The percentage of complete answers across your tracked prompts and engines that name you. Every prompt is asked once per engine per scan; an answer either names you or it doesn't. The delta compares that figure with the scan day before.

Why does my score change between scans? +

AI answers vary run to run and the sources an engine leans on can shift overnight. A new scan can name different brands even for the same prompt asked the same way. That's exactly why Referenced re-asks daily instead of checking once and assuming it still holds.

Why don't prompts with my brand name in them count the same way? +

If a prompt already says your brand name, asking whether you were mentioned isn't a meaningful test: the answer would be led by the question, not earned. That's why prompts are tagged by intent (branded vs unbranded) rather than blended into one number: it keeps a search you were always going to win separate from one you earned.

What is position? +

Where you're named among the brands the answer lists: first named is the top of the shortlist, not just any mention. AI answers list a small number of options and the order signals a preference, so being named third carries less weight than being named first.

Which engines count towards the score? +

Every engine switched on for your account. ChatGPT, Google AI Overviews and Perplexity run on every plan; Claude, Gemini, Grok, Meta Spark and AI Mode are optional add-ons. Turn one off and it stops feeding the blended score from that scan onward.

What's a good AI visibility score? +

There's no universal benchmark: it depends how many credible answers exist for your category and how many competitors are fighting for the same ones. The number that matters is your own trend against your own tracked competitors, not a fixed target.

See your visibility score before your competitors see theirs.

Free 7-day trial, no card: the first scan runs on all prompts across three engines.

Feature

Visibility

Your score across every AI engine.

Feature

Sentiment

What AI thinks of your brand.

Feature

Competitors

Share of voice, brand by brand.

Feature

Citations

Which pages AI actually cites.

Feature

AI Traffic

GA4 + Search Console, overlaid.

Feature

Opportunities

The fix for every gap, ranked.