MentionFlow AI
Open app
DocsMetrics

Visibility

A 0 to 100 score for how prominently AI engines mention your brand, weighted so an earlier mention counts for more than a later one.

Visibility is MentionFlow's headline score. It measures how prominently engines mention you, not just whether they do. Being named first is worth more than being named eighth.

How it's calculated

For every run in the window, we take the position of your first mention. That is 1 for first-named, 2 for second, and so on. Each position gets a rank weight of 1 / log2(1 + position). Runs where you don't appear get a weight of 0. The score is the average weight across all runs, scaled to 0 to 100:

visibility = 100 × Σ (1 / log₂(1 + position)) / (total runs in window)
  • A run where you weren't mentioned still counts in the denominator. Absence lowers the score, as it should.
  • 100 means you were mentioned first in every sampled answer.
  • 0 means you were never mentioned.

The colour bands, and where they come from

Visibility is the first of the five headline cards on your Overview. The score is painted red, amber or green, and so is the card's top hairline. The band word sits on the line beneath it. The cutoffs come from the score's own formula:

BandScoreWhat the formula says
Low (red)under 20Mostly absent, or named late in a small minority of answers. Roughly a one-in-five first-place equivalent or less.
Building (amber)20 – 50Present but inconsistent: named first in about a fifth to a half of answers, or mid-pack in most. Real, not yet reliable.
Strong (green)50 and upThe mean answer names you around third or better. 1 / log₂(1 + 3) is exactly 0.50, which is where the 50 comes from.
  • They are anchored, not absolute. 50 is a real point on the position-weight curve, so the boundary means something you can check. It is still not a target. A category with three plausible answers has a much higher ceiling than one with thirty.
  • The colour is never the reading. The hue and the number come from the same score, and the band word spells out the verdict right beneath it. Hover the band word to see the basis it was cut against.

Why it matters

  • It is the single number to watch over time. Because it is position-weighted, it rewards work that moves you up an answer, not just getting listed near the bottom.
  • It compares cleanly across prompts and competitors. Every entity is scored the same way over the same runs.
Note

Visibility blends only the search-grounded engines on headline surfaces. Per-engine views show each engine unblended. See Data honesty.

How to read it

  • Judge the trend, not the day-to-day ticks. Engine answers are non-deterministic, so small movement inside a window is expected noise.
  • Below five runs it carries a low-confidence flag. Give it more collection days before drawing conclusions.

Related