# Why numbers differ between screens

> Why Home, Topics, Competitors, MCP and Looker Studio show different AI visibility numbers for the same brand, and how to compare them fairly.

Source: https://getintel.ai/docs/academy/metrics/why-numbers-differ/

If Home says one visibility number and Topics another, both can be right. Each screen answers a slightly different question, over a different window or a different set of questions.

## What each number covers

| Where | What it covers | Window |
| --- | --- | --- |
| **Home** | Answers to your topics' buyer questions | The period you pick (7 days by default). The daily chart plots each day's own answers. |
| **Topics (Found)** | One topic's buyer questions | Rolling 7 days |
| **Competitors** | The same answers, per brand | The period you pick |
| **MCP `get_visibility`** | All tracked questions, brand-wide | Trailing 7 days, with 12 weeks of history |
| **Looker Studio** | One row per day, question and engine, summed or averaged however your chart groups them | The report's date range |

## The usual reasons for a difference

1. **Different windows.** Home over 30 days and a topic over 7 days describe different stretches of time.
2. **Different questions.** A topic covers only its own questions. `get_visibility` counts every tracked question, not only topic questions.
3. **Averaging.** Home averages across questions, so a topic with few questions carries the same weight per question as a big one. An average of topic scores is not the same as the overall number.
4. **Today is partial.** Until today's scan finishes, today's point on a daily chart rests on fewer answers.
5. **Grouping in [Looker Studio](/features/looker-studio/).** Adding Source Domain or Competitor Named to a chart repeats each answer once per domain or competitor. Keep those dimensions off headline totals.

<Aside type="tip">
To compare two screens, set them to the same period first. Most differences disappear.
</Aside>