> ## Documentation Index
> Fetch the complete documentation index at: https://docs.mavel.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Position

> Where in the answer your brand appears when it appears: the average rank across all mentions in the selected window.

Position measures how early in an answer your brand is mentioned, averaged across every mention in the selected window. It is expressed as a rank number: **#1.0** means your brand was always the first brand named; **#4.5** means it was named, on average, between the fourth and fifth brand in the answer.

Lower is better. A positive delta (rank number increased) means your average position worsened: you moved further down the answer. A negative delta means you moved up.

## How position is measured

When Mavel captures an answer, the extraction pipeline identifies every tracked-brand mention and records the ordinal position of that mention among all detected brand mentions in the answer (brands you don't track don't affect the rank). Those position values are summed and divided by the number of mentions with recorded positions to produce the average: `position_avg = sum(mention_positions) ÷ count(mentions with positions)`.

Position is only defined for answers where your brand appears. A prompt where you are absent does not affect your position average. It affects your [Visibility](/results/visibility) instead.

## Why it matters

AI answers have the same attention economics as search results: the opening recommendation carries the most weight. If a buyer asks "what tool should I use for X" and the answer spends three paragraphs on a competitor before a single sentence on you, that is a different situation from being named first with detail.

Position also signals narrative structure. Consistently late position often means AI engines treat your brand as a secondary or alternative option rather than a primary recommendation for that prompt category.

## How to use it

<Steps>
  <Step title="Check visibility first">
    Position is only meaningful where you appear. If visibility is low, fix that first: improving from "absent" to "mentioned late" raises visibility before it raises position.
  </Step>

  <Step title="Filter by topic and surface">
    Position varies by prompt category and surface. You may rank first on one AI surface and fifth on another for the same prompt set. The filter bar lets you isolate by surface, topic, or tag.
  </Step>

  <Step title="Read the delta">
    The period-over-period delta on the Position column tells you whether your average rank improved or worsened this period versus the prior equal-length window. Because lower is better, a negative delta (moving toward #1) is the positive signal.
  </Step>

  <Step title="Compare against competitors">
    The Top Brands table shows position for every tracked brand. A competitor ranking #1.5 while you average #4.0 on the same topic means they are structurally preferred for those prompts.
  </Step>
</Steps>

## How to interpret it

**Position #1.0-2.0** means your brand is consistently among the first mentioned: the most favorable structural position.

**Position #3.0-5.0** is middle-of-the-answer placement. You are present but not leading the recommendation. Source improvements (creating content that matches the citations the AI engine trusts) are the lever here.

**Position showing no value** has two possible causes: the brand never appeared in the selected window at all (check its Visibility), or it appeared but position ranks were not recorded for those mentions, which can happen for answers captured before the position pipeline was active for a given surface. The Visibility column distinguishes the two cases.

<Note>
  The delta display is inverted in the UI compared to Visibility and Sentiment: a green delta on Position means the number went down (improved), and a red delta means it went up (worsened). This matches the "lower is better" semantics of a rank.
</Note>

## Relationship to the other metrics

Position only tells you where, not how favorably. A brand mentioned first in a dismissive context still scores #1 for position. [Sentiment](/results/sentiment) is the companion metric that tells you whether early mentions are positive or negative. Together, position and sentiment describe the quality of your AI presence, not just its existence.