Metrics & ScoresKey Term

AI Reputation Index (ARI)

Also known as: ARI, AI reputation score

The AI Reputation Index (ARI) is a 0-100 score of how AI models describe and recommend a brand, combining sentiment, recommendation share, source quality, consistency, and stability.

Vollständige Erklärung

The AI Reputation Index (ARI) is a 0-100 composite score of how AI models describe and recommend a brand, used across the Rankfor Index 2026. It weighs five components: sentiment (25%), recommendation share in buyer-intent prompts (25%), source quality of the citations grounding the answers (20%), consistency of the story across languages and phrasings (20%), and stability of the answer across repeated runs (10%). The Nordic-Baltic edition computes ARI from 35,640 grounded responses across 66 brands, 12 languages, and three AI models; the Poland edition applies the same scale to 46 consumer brands with Polish prompts. Scores map to named bands, and the 2026 leaderboard tops out at 67, which means every measured brand still has room to grow. ARI measures visibility and recommendability in AI answers; it does not rate company or product quality.

Fachbegriff

Weighted composite of normalized sentiment, recommendation, source, consistency, and stability scores

Forschungserkenntnisse

Our Rankfor Index 2026 Methodology research found: 35,640 grounded responses across 12 languages

Read the full research

Frequently asked questions

How is ARI different from the AI Visibility Score?

ARI is the reputation composite used in the Rankfor Index studies (published rankings of measured brands). The AI Visibility Score on the platform decomposes into four separately actionable components (PIS, SMR, ARD, PPMA) for ongoing brand monitoring. The Index brand pages map ARI onto the same 0-100 scale so the two read consistently.

What is a good ARI?

In the 2026 Nordic-Baltic edition, 60 and above put a brand in the leaders tier and the best composite was 67. In the Poland edition, leaders start around 70. The absolute number matters less than rank within category and movement over time.