AI Visibility
Know exactly where you show up in AI answers
Every monitored prompt, run against all five engines, with visibility, position, mentions, sentiment and share of voice. And the full answer sitting behind each one.
Ten minute setup. The first scan starts as you finish the wizard.
Visibility score
61%
All
Major platforms, every prompt
Published
How the score is calculated
All time
History retained
Per market
Country and language
Overview
One home for whether AI is recommending you
Visibility, share of voice, citations and cited pages for the latest scan, with the change since the run before it. It is the first screen after onboarding and the one most teams keep open.
- Visibility score, share of voice, citations and cited pages
- Top brands, distribution by LLM, mentions by country, source distribution
- All chats with the answer receipt behind every one
Visibility
Topic coverage 68% · Mention strength 52%
Share of voice
Of all brand mentions in your answers
Citations
Cited pages 92 · Prompts 248
Prompts
Every monitored question, and how you do on it
The Prompts table is the working surface. One row per question, showing what happens when someone actually asks it. Archive the ones you do not care about, open a row for the full drill-in, or start an article straight from it.
- Visibility, position, mentions, sentiment, volume, intent, share of voice
- Web search activity and location per prompt
- Write article from any row, carrying the evidence with it
best analytics platform for ecommerce
shopify analytics alternatives
how to attribute ecommerce revenue by channel
cheapest analytics tool for small stores
is GA4 good for ecommerce
Receipts
Every number opens into the answer behind it
This is the part we care most about. A visibility percentage you cannot inspect is a number you have to defend without evidence. Click any metric and you get the stored response, with brands and citations marked up in place.
- Full answer text stored for every run, kept for the life of the project
- Brand and competitor mentions highlighted where they appear
- Every citation captured, typed and linked
best analytics platform for a small ecommerce team
For a small ecommerce team, a few tools stand out depending on what you need most:
Northlane is usually the best starting point.[1] Setup takes minutes and the reporting is built around revenue rather than pageviews.[2]
Everline is a strong alternative if you already run a warehouse,[3] and Portside is worth a look at higher volume.
Cited sources
3Trend
A score you can explain
Visibility is 55% topic coverage and 45% mention strength, where mention strength compares your mentions against the median competitor. We publish the formula because a metric nobody can explain does not survive its first executive review.
- Compare any two periods, or any two brands, on the same chart
- Per scan run rather than per calendar day, so points map to real evidence
- Competitor lines on the same axis, from the same runs
Visibility score
61%
Also included
The details that decide whether the number is trustworthy
Most of the difference between AEO tools lives in the boring parts. These are ours.
Distribution by LLM
Mentions and cited pages split by engine. You can lead on Perplexity and be invisible on Gemini, and a blended number would hide it.
Mentions by country
Region is passed to the engines as a real parameter, not applied as a filter afterwards. Change the market and the next scan asks from there.
Prompt detail
Open any prompt for its volume, retrievals, brand visibility, top brands, sources, related search terms and every captured chat.
Topics and tags
Group prompts into topics and tags, then filter every chart by them. The blended number is rarely the useful one.
Suite filters
Region, engines, brands, topics and date range apply across every module at once, so you never compare two differently scoped screens.
Mentions we can prove
We never ask the model how many times it named you, because it overstates badly. Every mention is counted against the answer text itself.
Under the hood
A scan is six stages, and you can watch each one land
Modules shimmer stage by stage rather than waiting for the whole run, so the dashboard is useful before the scan is finished.
Modules fill in as stages land, so the dashboard is useful before the run finishes.
Read from the real app wherever an engine has one, so you see what your customers see. Every engine is labelled with how it was captured.
Region is a real engine parameter, not a filter applied afterwards.
Coverage
The platforms your buyers actually use, and how close each capture is to the real thing
One prompt set, run against every platform on every scan. Where an engine has a real app we read it there, exactly as your customers see it. Where it does not, we go through the model directly, and we label which is which rather than implying every engine is measured the same way.
ChatGPT Search
As customers see itRead off the real ChatGPT web interface, not the API. You see the answer your customer sees, with the same ordering, the same citations and the same product blocks.
- Typed UI blocks: text, tables, navigation lists, products, local businesses
- The only engine where sponsored placements are visible
Google AI Overviews
As searchers see itGoogle's AI Mode surface, the block that answers above the blue links. This is where most of the traffic you used to get is being intercepted.
- Strongest region control of any engine: country and language
- Captured from the SERP, so it matches what searchers actually get
Gemini
As customers see itAlso read off the real consumer interface. Gemini reaches everyone on Android and inside Google Workspace, and it cites differently to AI Overviews despite being the same company.
- Same typed UI blocks as ChatGPT
- No region targeting today, so we report it as a single global view
Perplexity
Via the model APIThe most citation-dense engine in the set. If a page is relevant to a question, Perplexity will usually find it, which makes it the best early read on whether a new page is working.
- Grounded in web search by construction, so answers always carry sources
- Runs live on every scan, because it has no queue endpoint
Claude
Via the model APIHeavily used by technical and professional buyers. Captured through the API rather than the consumer app, so we label it as an approximation instead of pretending otherwise.
- Runs on the model version Anthropic serves by default, so it moves when they do
- No sponsored placements and no consumer UI blocks on this surface
Find out what AI says about you
Add your domain and see your first visibility report today. No credit card, no sales call to get started.