Sentiment Score

The Sentiment Score analyzes what AI actually says about your brand — not just that you're mentioned, but how you're described. It captures the tone, claims, and framing that AI models use when recommending (or warning against) your product.

How Sentiment Works

For every AI response that mentions your brand, we extract the specific rationale — the text explaining why AI recommends you or what caveats it mentions. Our proprietary algorithm then classifies each claim as positive ("reliable uptime," "excellent for beginners"), neutral ("mid-range pricing," "founded in 2012"), or negative ("limited features," "expensive renewal prices").

The Sentiment Score is a composite number from 0 to 100. A score of 65 means the overall tone is moderately positive. Above 70 is strong positive sentiment. Below 40 indicates significant negative framing.

Highlights and Concerns

Beyond the overall score, Rankry extracts and groups specific claims into Highlights (what AI praises about you) and Concerns (what AI flags as drawbacks). These are the exact phrases and themes AI models use, aggregated across all responses. For example, if three different models all mention "limited integrations" as a drawback, that shows up as a top concern with frequency data.

Why Sentiment Matters

Two brands can have the same visibility but completely different sentiment. Brand A might be consistently described as "affordable but basic" while Brand B is positioned as "premium and reliable." Sentiment shapes the buyer's perception before they even visit your website. If AI consistently frames you with a particular narrative — positive or negative — that narrative influences purchasing decisions.

Sentiment by Model

Different AI models can have very different opinions about your brand. ChatGPT might praise your pricing while Claude flags your feature gaps. The per-model sentiment breakdown helps you understand where your reputation is strongest and where it needs work.

Updated on: 02/03/2026