How visible are you this week?
Share of voice — the field
The field →Act on this
Prioritised moves from your latest run
Mention rate by engine
whiskers = 95% CI · latest runProducts
| Product | Status | GEO score | Mention | Share of voice |
|---|---|---|---|---|
| Loading… | ||||
AI crawler activity
Which AI bots fetch your pages · last 30 daysWhat the engines say about you
Every mention, citation and recommendation across ChatGPT, Claude, Gemini & Perplexity — each number carries a 95% confidence interval so you can tell real movement from noise.
Share of voice across runs
Shaded band = 95% CI (overlapping bands ⇒ within measurement noise) · large dots = change beyond noise · ⚑ = recorded content change (see Experiments).
Mention rate by engine
Whiskers = 95% CI
Web-grounded vs parametric
Live web search vs the model's memory
By buyer intent
Engine × intent — mention rate
Find the engine / journey-stage combos where you're invisible
Claims that contradict your facts
Answer explorer
Every captured answer this run — filter it, then read the full transcript with your brand highlighted.
Measured lift from changes
Baseline run → content change → treatment run. ✓ = beyond noise · ~ = within noise
Portfolio & brand
Every tracked page and its brand-level roll-up — visibility across your whole catalogue, from the latest run of each.
All products
Latest run per product · click a row for the breakdown, or act from the last column| Product | Brand | Mention | Trend | SOV | GEO | Status | Next step |
|---|---|---|---|---|---|---|---|
| Loading… | |||||||
What to fix next
Everything we found, in one queue, ranked by impact. Mark something fixed and we re-check it — nothing closes until a fresh check says the finding is gone.
The queue
Verified fixed
re-checked and confirmed — proof the work landedWrong facts about you, out there
Engines quote these pages when they answer questions about you. We draft the correction; you send it. The most influential pages come first.
Site audit
Two separate questions: can answer engines reach your site, and is what they find worth citing. Findings from both land in Actions as tracked work.
Crawl and audit
Site-level issues
one row per check, not per pagePages audited
worst first| Page | Score | Failing | Fix pack |
|---|
Can AI crawlers reach you?
robots.txt per crawler + live user-agent testsCrawler by crawler
The field around you
Who the engines name in your category, how often, and where the gaps are.
Share of voice across runs — the field
who's rising, who's flatRival leaderboard
| Brand | Share of voice | Mention | Recommend | Next step |
|---|---|---|---|---|
| Loading… | ||||
Head-to-head — prompts they win
from this run's captured answersAnswers that name a rival but not you — each one is a concrete prompt to counter with content.
Where the engines get their facts
The domains cited when models answer about your category — your off-site corroboration footprint, and the gaps where only rivals appear.
Most-cited domains
Measurement history
Every retained run for this project since your baseline (older runs are pruned per your retention setting). Open one to read it in Answers.
| Run | Scope | Mention · Δ | SOV · Δ | Cost | Status |
|---|---|---|---|---|---|
| Loading… | |||||
Author pages AI can cite
Turn product facts into structured, server-rendered page families with JSON-LD, answer-first content and a 0–100 GEO quality score.
Recent drafts
Create a page family
Toggle sections to see the projected GEO score move. Publish-ready at 90.
Projected GEO score
0 sectionsDissect a live URL
Audit any published page for GEO quality and pull its facts into your draft product brief.
Runs your page through the Studio audit engine and hands off the result to Content Studio for editing.
GEO quality score
ExampleFindings
2 blockers · 2 warnings · 2 passingExtracted facts
Detected on the page — review before importing
Structured data
JSON-LD types found on the page
Extract from documents
Upload PDFs, docs or spreadsheets — every fact lands as a review-flagged draft before it enters your product brief.
Extracted facts
review before publishExample output — parsed facts land here
ICP Discovery
Give the AI your product and it finds your ideal customer profiles — reuse them as “asking as” viewpoints in Prompt Lab and Measurement.
Uses the facts saved for that product in Brand profile.
Prompts
The exact buyer questions every run asks each engine. Curate this set and you change what gets measured — nothing here is asked until you track it.Product brief
One organised set of facts about your product — what it does, who it's for, its use cases, FAQs and specs. Every fact stays a draft until you approve it.
Export to CMS
Ship published pages as static HTML or push straight to your connected CMS.
Static HTML
Self-contained pages with inlined JSON-LD. Drop into any host or CDN.
Push to CMS
Publish straight to your connected site as structured nodes.
Pages to export
0 selectedJSON-LD payload
ProductThe GEO library
Field-tested guides on getting your brand mentioned, recommended and cited accurately by AI assistants — how to plan, produce, publish and measure your way into AI answers.
Plan & usage
Track product and measurement usage against your plan limits.
Brand profile
Get started
Derived setup
Workspace settings
Manage this workspace, the engines you measure against, and your measurement credits.
Workspace
Team & access
Who can see and act on this workspace. Admins manage billing, keys, members and notifications; members run and read; viewers read only.
Notifications
Deliver drift, hallucination, budget and run-failure alerts to your channels — stop checking the dashboard to find out something moved.
What you track
Whether you optimise many content pages for one product (typical SaaS) or a catalogue of distinct products (e-commerce). This just renames things across the console.
Engines measured
Which AI engines every run queries
Measurement credits
Each run spends credits across your enabled engines — no API keys or setup required.
API keys
For programmatic access and forwarding web/CDN logs to POST /v1/crawler/ingest. Secrets are shown once at creation.
Feedback inbox
What your team is telling you from inside the app — newest first. Admins only.
Advanced · bring your own keys
Optional. Use your own provider account instead of credits.
Managed
When set, runs bill to your provider account and don't spend credits — for teams with existing OpenAI or Anthropic contracts. Keys are managed by the platform API.