Features
Everything you need to measure, win and prove AI visibility
Track how AI engines talk about you and your competitors, see why they say it, act on a prioritized plan, and measure the result with statistics you can defend.
- 14-day free trial
- Cancel anytime
- Unlimited team members
- Setup in 5 minutes
AI visibility tracking
How often AI recommends you, with the margin of error on every number
AI answers change from one run to the next, so a single answer proves nothing. We ask every engine each prompt several times per run and report visibility as a range, so you know what is real and what is noise.
- 3 samples per prompt and engine on every run
- 95% Wilson confidence interval on every rate
- Changes flagged only when a two-proportion z-test says they are significant
- Break down by engine, prompt, tag and time period
AI visibility by engine · last 30 days
Competitor benchmarking
See who wins each prompt, and by how much
Add the competitors you care about and get share of voice, visibility and average position for every one of them, side by side. Competitors named in the prompt itself are excluded, so 'X alternatives' never inflates X.
- Share of voice across all tracked brands
- Prompt-level leaderboard: who AI names first
- Average position when mentioned
- Trend lines per competitor, run by run
Share of voice · 12 prompts × 5 engines
| Brand | Share of voice | Prompts led |
|---|---|---|
| Competitor A | 31% | 12 |
| Your brand | 27% | 9 |
| Competitor B | 22% | 6 |
| Competitor C | 12% | 2 |
| Others | 8% | 1 |
Citation intelligence
Know which websites AI trusts in your category
Every answer's sources are stored and grouped by domain and page. You see which sites shape the answers, whether those answers mention you, and where a single placement would move the needle.
- Top cited domains and URLs per engine
- Your own site's citation rate, with confidence interval
- 'Also mentions you' ratio for every third-party source
- Opportunities: high-share sources that leave you out
Most cited sources
- g2.comopportunity34%
- reddit.com29%
- yourbrand.comyour site19%
- techradar.comopportunity17%
- capterra.com12%
Sentiment & narrative
Read the exact words AI uses about you
Every mention is scored from −1 to +1 and kept with the sentence it came from. You can open any number down to the verbatim answer, so a negative trend always comes with the claim behind it.
- Five-level sentiment per mention, per engine
- The snippet that produced each score
- Negative or outdated claims surfaced in the action plan
- Full answers with brands highlighted
How AI describes your brand
“…widely recommended for its reliability and transparent pricing.”
“…another well-known option that covers the basics.”
“…capable, though some users report it can be expensive.”
Action plan
A prioritized to-do list, backed by evidence
After every run we turn the data into concrete actions, each with the numbers that justify it. No generic advice: every item points at a prompt, a source or a line in your robots.txt.
- Content gaps: prompts where a competitor is recommended and you aren't
- Off-site: sources AI cites heavily that don't mention you
- Technical: blocked AI crawlers and a missing llms.txt
- Reputation: negative framing, with the quotes
Open actions
Win the answer for “best CRM for startups”
Competitor A in 78% of answers, you in 22%
Get featured on g2.com
Cited in 34% of answers; mentions you in 12% of them
Unblock AI crawlers in robots.txt
GPTBot and PerplexityBot are disallowed
Impact measurement
Prove that the work moved the numbers
Mark an action as done and we compare visibility on the prompts it targeted in the 28 days before and after, with a significance test. It is the before-and-after evidence that turns GEO from a cost into a result.
- 28-day windows before and after each action
- Scoped to the prompts the action targets
- Two-proportion z-test with the p-value shown
- Results flow into client reports automatically
Measured impact
Published a comparison page with a clear verdict
28 days before
31%
28 days after
54%
AI crawler audit
Make sure AI can actually read your site
If GPTBot or PerplexityBot are blocked, no amount of content will get you cited. We check your robots.txt against the crawlers that matter and look for an llms.txt every time we build your plan.
- Checks GPTBot, OAI-SearchBot, ChatGPT-User, PerplexityBot, ClaudeBot, Claude-SearchBot, Google-Extended and more
- Flags a missing llms.txt
- Shows up as a high-priority action when something is blocked
robots.txt · yourbrand.com
- GPTBotblocked
- OAI-SearchBotallowed
- ChatGPT-Userallowed
- PerplexityBotblocked
- ClaudeBotallowed
- Google-Extendedallowed
- /llms.txtmissing
White-label reports
Client-ready reports under your brand
On Agency plans every client gets a private live link with your logo and colors, and a one-click PDF. Visibility, competitors, completed work and its measured impact, explained in plain language.
- Your logo, name and brand color; our brand never appears
- Private share link per client, rotate it any time
- Save as PDF for monthly reporting
- Unlimited team members on every plan
Client report preview
Your agency
AI visibility report · Client
AI visibility
58%
+11.4 pts · significant
Work completed
6 actions
4 with measured lift
GEO writer
Rewrite the page AI should be citing
Pick a prompt you're losing. We read the answers AI gives, the pages it cites instead of yours and your own page, then tell you exactly why the others win and draft the content that closes the gap.
- Grounded in real answers and the pages AI actually cites, not generic SEO advice
- Section-by-section changes, missing facts, FAQ and JSON-LD, plus a full draft
- Off-site targets and technical fixes, added to your action plan in one click
Optimize · best eSIM for Japan
Why others win
- • Cited pages compare price per GB in a table; yours doesn't
- • They state coverage on named carriers (Docomo, SoftBank)
- • No FAQ answering "does it work on the bullet train?"
Draft · H2 "Japan eSIM plans compared"
5 plans · price per GB · carrier · hotspot…
AI traffic & crawlers
See the visits and bots AI sends your way
A 1 KB script attributes sessions from ChatGPT, Perplexity, Gemini, Copilot and Claude to the pages they land on. Upload server logs or connect your CDN to see which AI crawlers read which pages, and how often.
- AI referral sessions by engine and landing page, no cookies
- AI crawler hits (GPTBot, OAI-SearchBot, PerplexityBot, ClaudeBot…) per page
- GA4 CSV import for historical AI traffic
AI referrals · last 30 days
Reputation & alerts
Know when AI starts telling a different story
After every run we summarize how AI describes you: strengths, weaknesses and claims that may be outdated or wrong. Get an email, Slack message or webhook when visibility drops significantly, a competitor overtakes you or sentiment turns.
- Reputation score with the claims behind it, quoted from real answers
- Alerts only on statistically significant changes, so they mean something
- Email, Slack and signed webhooks
Alert · Slack
Visibility dropped on Perplexity
48% → 31% (−17 pts, p = 0.01) · Airalo now leads 6 of 20 prompts
Possibly outdated claim
"No unlimited data plans" · repeated in 4 ChatGPT answers
Answer signals
Everything inside the answer, not just the text
We keep what engines show around the answer: the searches they ran behind the scenes, ChatGPT shopping cards and ads, local business packs, and the Reddit threads and YouTube videos they lean on.
- Query fan-out: the sub-searches AI used to build its answer
- ChatGPT shopping products, sponsored placements and local packs
- Reddit and YouTube sources ranked by how often they're cited
Fan-out · best eSIM for Japan
- japan esim unlimited data
- esim docomo coverage 2026
- airalo vs holafly japan
- reddit best esim tokyo
AI analyst
Ask your data in plain language
“Where are we losing to competitors, and why?” The analyst reads your metrics, real answers, sources and action plan before it replies, and every number it gives you comes from your data.
- Answers with evidence: quoted excerpts, sources and significance
- Suggests concrete next steps from your own action plan
- In any language your team writes in
Ask the analyst
Did our visibility really change this month?
Yes on ChatGPT: 34% → 49% (p = 0.02), after the comparison page went live. On Gemini the change (+3 pts) isn't significant yet.
Power BI, Looker Studio & API
Your AI visibility in the reports you already send
Daily tables for visibility, prompts, sources, AI traffic, crawlers and reputation plug straight into Power BI, Looker Studio, Google Sheets and Excel, with scheduled refresh. Everything the app does is also available through a documented REST API and an MCP server.
- Looker Studio connector and ready-made Power Query for Power BI and Excel
- Read-only keys made for BI tools; one report across every client for agencies
- Full REST API with OpenAPI spec, MCP server, signed webhooks and CSV on every plan
Looker Studio · Huginnova connector
AI engines
The engines your buyers actually use
Answers are collected with web search on, in the country and language you choose. Every answer records how it was collected.
ChatGPT
All plansThe real ChatGPT interface, as a user sees it, with the sources it cites.
Perplexity
All plansPerplexity Sonar, which searches the web for every answer.
Gemini
All plansThe real Gemini interface, as a user sees it, with the sources it cites.
Google AI Overviews
All plansThe AI Overview shown on a live Google results page, via DataForSEO.
Google AI Mode
All plansLive Google AI Mode results, via DataForSEO.
Claude
Add-onClaude with web search on, the way people use it on claude.ai.
Find out what AI says about you
Set up your first brand in five minutes and get real answers from every major AI engine the same day.