Being named first outweighs being named often

50% — of each engine score is first-mention rate. The Crackle PR AI Visibility Index weights first-mention rate at 50% of each engine's score, ahead of mean rank inverted (25%) and total mention rate (25%).

Published by Parry Headrick, Crackle PR · Updated 2026-09-22 · CC BY 4.0 · Category: Method

TL;DR

Engine score = 50% first-mention rate + 25% mean rank inverted + 25% total mention rate.
Composite = simple average of the per-engine scores.
Weights are frozen in the public methodology spec and re-scorable from the raw CSV.

Why this number matters

Inside a generated answer there is no scroll and no second impression. The first company named frames the comparison that follows, and in shortlist-style answers it is frequently the only name a buyer carries into the next step.

Crackle PR made that judgment explicit rather than burying it in a black-box composite. Half of every engine score is first-mention rate, a quarter is inverted mean rank, and a quarter is total mention rate — published, frozen, and reproducible from the public CSV.

The weighting is a stated position, not a neutral fact, and it is falsifiable: anyone can re-score the same raw data under different weights and publish the difference. That is the point of shipping the dataset under CC BY 4.0.

We published the weights so someone can prove them wrong. A benchmark you cannot re-score is an advertisement.
— Parry Headrick, founder, Crackle PR

How it was measured

Weights are declared before each edition is run and frozen in the machine-readable methodology spec at /data/methodology.json. Composite scores are the unweighted average of the per-engine weighted scores.

Denominator

Every scored answer in the edition: 25 queries × 4 engines × 5 runs = 300 answers for Q3 2026; 625 from Q4 2026 with Grok added as a fifth engine.

What this number does not prove

The weighting encodes an editorial judgment about how answers are read. A reader who believes total mention rate matters more should re-score the published CSV under their own weights — Crackle PR's ordering is not guaranteed to survive that, and the license permits it.

Cite this statistic

Reuse is permitted under CC BY 4.0 with attribution. The underlying data is public: raw CSV, JSON, and the DOI deposit. Full scoring rules: Index methodology.

Attribution line: Source: Crackle PR AI Visibility Index (https://www.cracklepr.com/ai-visibility-index), Q3 2026 edition, CC BY 4.0.

Quotable sentence: The Crackle PR AI Visibility Index weights first-mention rate at 50% of each engine's score, ahead of mean rank inverted (25%) and total mention rate (25%).

Embed (HTML): <blockquote>The Crackle PR AI Visibility Index weights first-mention rate at 50% of each engine's score, ahead of mean rank inverted (25%) and total mention rate (25%). <cite><a href="https://www.cracklepr.com/stats/first-mention-rate-weighting">Crackle PR AI Visibility Index, Q3 2026</a></cite></blockquote>

Academic citation (APA): Crackle PR. (2026). Crackle PR AI Visibility Index (Q3 2026 edition) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.21986226

Go deeper

AI Visibility Index — The full leaderboard, dataset, and quarterly editions.
Index methodology — Query basket, scoring weights, and margin of error.
Publisher recusal policy — Why Crackle PR holds no rank in its own leaderboard.
First-mention rate — Definition of the Index's most heavily weighted metric.