Outrigger Help
AI Monitor: Track Your Brand's Visibility Across AI Providers
The AI Monitor tests whether OpenAI API diagnostics, Perplexity API diagnostics, Gemini API diagnostics, Anthropic API diagnostics, Google AI Overview collector (provisional) mention your brand when asked buying-intent questions, and tracks your share of model over time.
In this article
What AI Monitoring Measures & Why It Matters
AI Monitor answers the question every brand should be asking: "When someone asks ChatGPT about my industry, does it mention my brand?"
Traditional SEO tracks Google rankings. Outrigger's AI Monitor tracks AI visibility — whether your brand appears in supported provider responses when people ask buying-intent questions. The confirmation identifies the exact tests included before you approve a run.
This matters because AI-powered search is growing fast. Perplexity alone processes millions of queries daily, and ChatGPT's browsing feature sends traffic directly from AI conversations. If your brand isn't being recommended by these models, you're invisible to a growing segment of potential customers.
Key metrics the AI Monitor tracks: - Share of Model — The percentage of test prompts where your brand is mentioned - Mention Rate per Model — Which AI models mention you vs which don't - Recommendation Rate — Not just mentioned, but actively recommended - Link Citation Rate — How often AI models link to your content - Competitor Share — How your competitors compare on the same prompts - Trend Over Time — Whether your visibility is improving or declining
Setting Up Monitoring Prompts
Monitoring prompts are the questions used to test brand visibility. Enabling a keyword prepares its buying-intent prompts. Review the available tests and any supported schedule in the confirmation before approving. You can also write custom prompts for manual tests.
Auto-generated prompts use your keywords to create buying-intent questions: - "What are the best [keyword] options?" - "Can anyone recommend a good [keyword] service?" - "I need help with [keyword] — what do people suggest?" - "Compare the top [keyword] companies"
Custom prompts let you manually test specific scenarios: - "[Competitor] alternatives that are more affordable" - "Best [product category] for [specific use case]" - "Is [your brand] worth it compared to [competitor]?"
Tips for effective prompts: - Write prompts the way real people ask AI models (conversational, specific) - Include comparison prompts ("[Brand A] vs [Brand B]") - Test different intent levels: informational, commercial, transactional - Keep each question specific enough that a provider can give a useful buying recommendation - Custom prompts are manual-only. Scheduled checks use the generated prompts for the keywords you separately enable and approve.
Reading the Timeline & Trend Data
The AI Monitor dashboard shows your visibility data in three views.
Overview — Your current share of model, broken down by AI model. A stacked bar chart shows which models mention you vs which mention competitors.
Timeline — A trend line showing your share of model over weeks/months. Event markers highlight when you placed citations or ran press campaigns, so you can correlate actions to visibility changes.
Results Detail — Individual test results showing the exact prompt, the AI model's full response, whether your brand was mentioned, the context of the mention, and which sources were cited.
What good looks like: - Share of model above 30% = strong visibility - Share of model 10-30% = growing, keep seeding - Share of model below 10% = early stage, focus on citation volume - Upward trend = visibility improved across the measured runs; it does not by itself prove which action caused the change - Flat line = no measured movement in this period; review the saved evidence before deciding what to change
Setting Up Alerts
Configure alerts to get notified when important changes happen to your AI visibility.
Available alert types: - Share of Model Drop — Alert when overall share of model falls by your chosen number of points - Share of Model Rise — Alert when overall share of model rises by your chosen number of points - Competitor Spike — Alert when a competitor's share of model jumps by your threshold - New Model Appearance — Alert when a provider starts mentioning the brand - Brand Lost — Alert when the brand stops appearing in a provider where it previously appeared - Response Changed — Alert when your chosen number of responses changes materially
Each rule shows its threshold and delivery choices before it is saved. Available destinations include the agency owner email, additional approved email addresses, a Slack incoming webhook, and the assistant inbox.
Pro tip: Set a "competitor surge" alert for your top 3 competitors. If a competitor suddenly jumps in AI visibility, investigate what they did — new press coverage, new forum posts, or updated content — and respond quickly.
Frequently Asked Questions
Which AI models does the monitor test?
Outrigger supports OpenAI API diagnostics, Perplexity API diagnostics, Gemini API diagnostics, Anthropic API diagnostics, Google AI Overview collector (provisional). The confirmation shows the exact providers approved for a run, and the results distinguish providers that succeeded, failed, or were skipped.
What is share-of-model?
Share of model measures what percentage of completed, relevant provider tests mention or recommend your brand. It is like share of voice for AI recommendations, with failed or skipped providers shown separately rather than counted as successful responses.
How often should I test?
Use the cadence and first eligible date shown in the monitoring confirmation. Scheduled checks use the approved keyword prompts; custom prompts remain manual unless a separate recurring workflow is offered and approved.
Can I track competitor visibility too?
Yes. The AI Monitor automatically extracts competitor mentions from every test response. Over time, you can see how your share of model compares to competitors and whether your citation seeding is closing the gap.