Help centerMonitor
Insights
Trends, brand standings, the performance map and what AI says about you.
Insights compares your brand with competitors over time. It shows where you win or lose by model and topic, and what AI actually says about each brand. Every panel uses the date, model, topic and tag filters at the top.
The header
The headline sums up the range, for example “AI recommends Acme in 42% of answers that suggest tools”. Below it, four tiles each show the change since the previous period:
- Visibility
- What it tells you: Share of answers that name you (“Named in N of M answers”).
- Shows a number after: 10 answers
- Recommended
- What it tells you: Of the answers that recommend any product, the share that recommend you.
- Shows a number after: 5 answers that suggest tools
- Position
- What it tells you: Your average place in the answer, and how often you are named first.
- Shows a number after: 3 mentions
- Sentiment
- What it tells you: 0–100. 60 or more is Positive, 40 or less is Negative, and in between is Neutral.
- Shows a number after: 3 mentions
| Tile | What it tells you | Shows a number after |
|---|---|---|
| Visibility | Share of answers that name you (“Named in N of M answers”). | 10 answers |
| Recommended | Of the answers that recommend any product, the share that recommend you. | 5 answers that suggest tools |
| Position | Your average place in the answer, and how often you are named first. | 3 mentions |
| Sentiment | 0–100. 60 or more is Positive, 40 or less is Negative, and in between is Neutral. | 3 mentions |
If one of your topics rarely gets product answers, a notice says “AI rarely names products for …”. When fewer than a quarter of a topic’s answers name any product, its brand numbers say little. Review promptstakes you to the Prompts page to rephrase them.
Brand trend and Brand standings
The chart and the table beside it always show the same metric. Pick one with the tabs:
- Visibility
- Share of answers that name the brand.
- Recommended
- Share of product-recommending answers that recommend the brand. Days with answers not read yet are shaded Not read yet.
- Sentiment
- 0–100, where 50 is neutral.
- Position
- Average rank. The axis is flipped so that #1 is at the top.
- Share of voice
- Each mention counts as 1 ÷ its rank, so being named first weighs more than being sixth in a list. A brand’s share is its weight divided by the weight of all tracked brands.
- The chart starts with the top five brands for the metric, always including yours (the thick blue line). Use D W M to group the points by day, week or month.
- Brand standings ranks every tracked brand. Click a row to add it to the chart or remove it; the chart holds up to 8 brands.
- A gap in a line means that period had too few answers for a fair number. It is not a drop to zero.
What AI says
This panel shows how answers describe a brand. Pick the brand with the Brand menu; yours is listed first.
- Role bar
- The share of answers that recommend, compare, merely mention or warn against the brand.
- Praised for / Criticized for
- The strengths and weaknesses answers mention most, with how many answers say each one. Click a point to list those answers.
- Quotes
- Real sentences from the answers, both positive and negative, with the model and prompt. Click one to open the whole answer in Answer audit.
AI also names
Brands AI names in at least two answers that you don’t track yet. Named is the share of read answers that name the brand. Recommends is how often those answers recommend it.
- Track adds the brand as a competitor. Past answers that name it are read again, so its history fills in.
- × (Not a competitor) hides it so it is no longer suggested.
Only owners, admins and editors see these buttons. See Brands for everything about competitors.
Performance map
A grid of AI models (rows) by topics (columns) for the metric you choose in the map’s own tabs. Each cell shows your value and how many answers it uses. A pink tint means a competitor leads there. Topics are sorted by the largest gap. The first 6 are shown; Show all N topics reveals the rest.
Select a cell
The largest gap is selected when the page opens. Click any cell to choose another.Compare
The panel below shows your value against the leading competitor in that cell, and the gap between you.See the answers
Under Answers with this gap is each answer behind the gap, with the sites it cites. Click one to read it. Answer audit ↗ opens every answer for that model and topic.Act on it
Use Add recommendation to save an action with the gap answers attached as evidence. You can edit the title, description, priority and effort first.
What changed
When a previous period exists, this list shows the two model–topic cells that gained the most and the two that lost the most. Each one links to its answers.
Reading the markers
Chart markers
- The AI models changed on this day (hover for which)
- Answers on these days are not read yet
- Your brand
- A competitor
After a model change, change markers compare both periods using only the models that answered in both. The marker’s tooltip lists them (“same models: …”). See Filters.
Related guides
- Metrics explainedEvery number in GEO Radar: what it counts, how it is calculated and when it shows a dash.
- BrandsYour brand and competitors: names, domains, colours, matching and suggested competitors.
- RecommendationsSuggested actions from your data: accept, plan, move through statuses and verify fixes.