Outrigger Help
What Outrigger can and can't tell you
How Outrigger samples AI visibility, which AI assistants the audit and Monitor check, why scores can change between runs, what remains outside our view, and what stays under human review.
In this article
Scores Are Sampled Measurements
Outrigger scores are sampled measurements, not a permanent census of every AI conversation. AI answers vary by prompt wording, time of day, user location, and model updates — so re-running an audit or monitor check can give a different result even when your website has not changed.
Use your score as a directional baseline across a consistent prompt set rather than an exact, fixed rank.
Which AI Assistants Audits and Monitor Check
The AI visibility audit (including the free audit) checks four AI assistants: ChatGPT, Claude, Gemini, and Perplexity.
Ongoing monitoring (AI Monitor / Share of Model) covers five AI experiences: ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews.
Outrigger does not track other assistants that are not listed above, and when a provider check fails or is unavailable during a run, Outrigger marks that check separately rather than counting it as a zero.
Audits Take Time to Run
Audits query live AI providers and third-party sources, so they take time to run rather than returning right away. Progress updates appear on screen as each pillar finishes its checks.
What Outrigger Cannot See
Outrigger only observes the public sources and sampled prompts it runs. There are several things we cannot see:
- Private or logged-in AI answers — We cannot see private user chats, personalized memory, or workspace-specific enterprise AI conversations.
- Every possible prompt — Real buyers phrase questions in countless ways. Outrigger tests a sampled set of prompts for your approved keywords, not every question a person might ask.
- Why a competitor appears inside a model's weights — Outrigger can show which public citations, reviews, entity profiles, and URLs appear alongside a recommendation, but no external tool can inspect a closed model's internal reasoning to prove the exact cause of a single answer.
Work Is Drafted for Human Review, Not Sent Automatically
Outrigger prepares research and drafts so your team stays in control of what goes live:
- Community responses and review replies are drafted in your workspace for you to review, edit, and post yourself.
- Website schema and technical recommendations are generated as code or instructions for your team to review and deploy on your site.
- Press releases and journalist pitches are drafted in Outrigger and only sent after an authorized user reviews and approves the outreach.
Frequently Asked Questions
Why did my score change when I ran another audit?
Outrigger scores are sampled measurements. AI answers vary by prompt wording, time, location, and provider model updates, so running the same check again can produce a different result.
Which AI assistants does Outrigger check?
The AI visibility audit (including the free audit) tests four AI assistants: ChatGPT, Claude, Gemini, and Perplexity. Ongoing monitoring and Share of Model cover five: ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews.
Does Outrigger post comments or send outreach automatically?
No. Forum replies, review responses, schema updates, and press pitches are drafted for a human to review, edit, and publish or send.