One loop: measure, fix, prove.
Eight capabilities, one workflow. Everything below is included in Pro — $9.99/month.
Answer Insights
See when AI mentions you, when it actually recommends you, and which competitors it names instead — across ChatGPT, Google AI Overviews, Perplexity, and Claude, broken down by the kind of question asked.
How it works
We build a set of questions your real buyers ask (branded, category, comparison, problem), run each one many times per engine, and read every answer for mentions, recommendations, citations, and sentiment.
Why it's different
AI gives different answers each time you ask. Every number we show carries a confidence range and a run count, so you always know what's signal and what's noise.
Demand Intelligence
Know which questions are worth winning. Each prompt gets a demand tier — high, medium, low — so you spend effort where buyers actually are.
How it works
Tiers come from public signals (Google Autocomplete, community discussion frequency, and real search volumes when connected) and from your own Search Console queries and sales-call questions, imported in one paste.
Why it's different
We never invent a fake 'AI prompt volume' number. Tiers state their derivation, and your own funnel data — the one perfectly unbiased sample — gets first-class treatment.
Experiments
The feature the monitors don't have: proof. Ship a fix and find out whether it actually moved your visibility — or did nothing.
How it works
Pick a fix from your prescription board. We split the matching prompts into a treatment group and a held-out control, re-measure after you ship, and run a significance test.
Why it's different
Verdicts are honest: worked, noise, or kill. When we say a fix worked, it's because it beat the control — not because a dashboard line wiggled.
Monitoring & Alerts
Your visibility, re-checked daily or weekly, with alerts when something real changes — a drop after a model update, a surge after a placement lands.
How it works
Scheduled re-measurement with the same rigor as your baseline, compared batch over batch.
Why it's different
Alerts fire only when a move exceeds the statistical noise band. No 3am panic over sampling variance.
Page Audits
Point at any URL and get a pass/fail checklist of the on-page factors proven to affect AI visibility — plus whether AI crawlers can even read your site.
How it works
We fetch the page and score it: does it open with a direct answer? concrete statistics? credible citations? a visible date? hedging? Then we check llms.txt and robots.txt for AI-crawler access.
Why it's different
Each check carries the strength of evidence behind it — and we tell you which popular 'optimizations' (schema stuffing, visual formatting) measurably do nothing.
Agent Traffic
See which AI crawlers — GPTBot, ClaudeBot, PerplexityBot, Google-Extended — are actually reading your site, and what they read.
How it works
Paste your server access logs; we detect every known AI crawler and summarize requests and top pages.
Why it's different
Parsing runs entirely in your browser. Your logs never leave your machine.
Rewrite Assistant
Turn a weak paragraph into the passage an AI answer-writer wants to quote.
How it works
Paste one passage; it comes back front-loaded, specific, de-hedged, and citation-ready — with a note on exactly which levers were applied and why.
Why it's different
One passage at a time, deliberately. This is a scalpel for your money pages, not a content farm.
Client Reports
A print-ready report — diagnosis, actions, proof of lift — branded with your agency's name and colors.
How it works
One click from any project; your byline and accent color come from settings. Print → Save as PDF is the deliverable.
Why it's different
The methodology footnote is built in, so your clients can trust the numbers the same way you do.