AI monitor reference

Prompt fields

Field	Meaning	Values
Prompt text	The question sent to the AI assistants	Up to 1000 characters
Status	Whether the prompt is asked	Active, Suggested, Archived
Source	How the prompt was created	Manual, Suggested, Imported
Region	Market the answer is collected for	See regions below; empty means the project market
Tags	Project tags for grouping and filtering	Any project tag
Personas	Personas linked to the prompt	Any project persona; the row shows two, then “+N”

Only Active prompts are asked. Editing a prompt’s text, status, or region is possible in the app; personas and tags are set for you by your account manager.

Regions

France, Germany, Belgium, Luxembourg, Netherlands, United Kingdom, United States, Flanders, Wallonia, Brussels.

Selecting your project’s own market in a region filter also returns prompts that have no region of their own.

Filters per tab

Tab	Filters
Overview	Date range, Tags, Models
Prompts	Search prompts, Status, Region, Tags, Personas
Answers	Date range, Tags, Models, region, sentiment, Only

Date-range options: week, month, 3 months, 6 months, year, all, custom. The default is month.

Metric definitions

Metric	Definition
Visibility (KPI)	Percentage of answers in the date range that mention any of your own brands
Visibility (Rankings)	Each brand’s share of all brand mentions in the period; the column sums to 100%
AI share of voice	Your brand’s slice of the total brand-mention pool across all monitored brands; each answer counts at most once per brand
Position	Average ordinal place of the brand in the answers that mention it; 1 = mentioned first
Average sentiment	Average sentiment score of the answers in the range, from −1 to +1
Sentiment label	Above 0.3 = Positive; below −0.3 = Negative; otherwise Neutral
Top domains	Domains cited in answers; the count is the number of distinct answers referencing the domain, the percentage its share of all answers in the range
Trend badge	Comparison with the previous range of the same length

Brand mentions come from the automatic analysis of each answer, not from raw text matching.

AI assistants and models

One model per monitored assistant, chosen to match what that assistant serves its own users by default:

Assistant	Provider
ChatGPT	OpenAI
Gemini	Google
Claude	Anthropic

GripIQ keeps this in step with the assistants themselves: when an assistant changes the model it serves its users by default, GripIQ moves to that model with it. Answers are grouped per assistant, so a model update never splits your filters or metrics — you may only notice a shift in an assistant’s answers or trend line around the date it changed.

The Models filter lists only providers that actually have answers in your project, and appears only when there is more than one.

Schedule

Interval	Prompts are asked
Daily	Once a day
Weekly	Once a week
Monthly	Once a month
Disabled	Never

The interval is a project setting. There is no button to run a prompt on demand.

Web search and sources

When answering, the assistant may search the web, biased to the prompt’s region and otherwise to the project market. The pages it consulted are stored per answer and shown under Sources. The Top domains card is built from those citations.

Exports

Export	Where	Columns
Answers (filtered set)	Answers tab → Export CSV	Prompt, Answer, Model, Brands Mentioned, Sentiment, Date
Answers for one prompt	Prompt detail → Export CSV	Same

The file is named ai-monitor-answers-YYYY-MM-DD.csv.

Limits

• Prompt text: 1000 characters.
• Additional instructions in the generate dialog: 255 characters.
• Count in the generate dialog: 1 to 10 prompts per run.
• Recent answers on the prompt detail page: 15 per page.
• Top domains: the top 10 domains.
• Rankings: 5 brands per page.

The number of prompts your project may track is set by your plan. See Plans.