Monitors

Monitors define which AI models to test and how often. They're the engine behind Cleotic's data collection -- each monitor runs your prompts against the AI models you've selected on the schedule you've configured.

Why monitors matter

Different AI models can have very different views of your brand. ChatGPT might mention you frequently while Claude barely knows you exist. A monitor lets you configure exactly which models to test and at what frequency, so you get a complete picture across the AI landscape.

You can also use multiple monitors within a project to segment your monitoring -- for example, one monitor for the US market and another for the EU, or one for daily quick checks and another for comprehensive weekly analysis.

Creating a monitor

You can create monitors during the project wizard (Step 3) or from the Monitors tab on your project page.

When creating a monitor, you configure:

Name

A descriptive label for this monitor (e.g., "Daily US English", "Weekly EU Spanish").

Geographic context

The market or region this monitor is focused on (e.g., "US", "UK", "EU", "Global"). This is a label for your own organisation -- it helps you remember which monitor covers which market.

Language

The language context for this monitor (e.g., "English", "Spanish", "German"). Like geographic context, this is an organisational label.

AI models

Select which AI models to query. Models are grouped by provider:

• OpenAI -- GPT models (consumer and pro tiers)
• Anthropic -- Claude models
• Google -- Gemini models
• And others as available

Each model uses a vendor:tier format internally (e.g., openai:consumer, anthropic:pro), but you'll see friendly display names in the selection interface.

Select at least one model. More models give you broader coverage, but each additional model means more API calls per prompt run.

Refresh frequency

How often your prompts run automatically:

• Daily -- Good balance for most use cases
• Weekly -- Suitable for long-term trend tracking with lower usage

Web search

When enabled, AI models can use web search to supplement their responses. This can surface more current information and more citations, but is only available on Custom and Grandfathered plans.

Monitor status

Like projects, monitors can be active or paused:

• Active -- Prompts run on the configured schedule
• Paused -- Scheduled runs stop, but existing data remains accessible

Pausing a monitor is useful when you want to temporarily reduce usage or focus on specific monitors. You can pause and resume from the monitor detail page or the monitors list.

The monitor page

When you open a monitor, you see:

• Stats grid -- Geographic context, language, number of models, refresh frequency, and prompt limit at a glance
• Prompts tab -- All prompts in this monitor, with options to add, run, and delete. See Prompts.
• Analytics tab -- Visibility scores, share of voice, and model breakdown specific to this monitor

Tips

• Start with daily frequency. It gives you enough data to spot trends without excessive API usage.
• Use multiple monitors for different segments. A "US English" monitor and a "UK English" monitor can reveal how AI visibility differs across markets.
• Don't over-select models. Each model adds to your prompt run time. Start with the most popular models (ChatGPT, Claude, Gemini) and add others as needed.

Plan limits

See Plans and usage for full details.