Skip to main content

​
Overview

Sentiment measures the tone of what AI says about your brand, on the runs where your brand is mentioned. It’s scored on a 0-100 scale — 0 strongly negative, 50 neutral, 100 strongly positive — reflecting how the AI framed you in context, not just whether the mention was positive or negative in isolation.

​
How it’s calculated

Sentiment is worth 10 of the AIS Score’s 100 points — the smallest weight of the four signals, because it only qualifies mentions that Visibility has already established exist. Sentiment is calculated only for runs that contain a brand mention, using an anchored 0-100 scale evaluated by a language model in context. Like Position, the raw when-mentioned average then passes through the same credibility shrink and ramp before it enters the composite:
With zero mentions, Sentiment is 0. A brand that’s mentioned but hasn’t had its sentiment scored yet reads as neutral (50) rather than as missing data — this can happen transiently between a mention being detected and the sentiment pass completing. Once real sentiment scores accumulate, the credited value moves away from that neutral placeholder in proportion to the evidence, exactly as Position does.
Sentiment and Position share the same shrink constant (K = 8) and the same 3-mention ramp — they’re both “when-present quality” signals, credibility-bound in the same way for the same reason. See Position for the full mechanics.

​
Why it matters

Sentiment tends to move less dramatically than Visibility or Position, and that’s expected — most AI answers describe brands neutrally-to-positively by default rather than taking a strong stance. A genuinely low Sentiment score is worth investigating directly: look at the actual answer text in your mentions feed or the dedicated sentiment monitoring view to see what’s driving the tone, since a single 0-100 number can’t tell you why an answer read negatively. A low Sentiment score paired with strong Visibility and Position is a specific, actionable pattern: you’re the name AI reaches for, but something in how you’re described is dragging the framing down — outdated information, a recurring complaint theme, or an unfavourable comparison are common causes worth chasing down in the underlying answers.

AIS Score

How Sentiment combines with the other three signals into the headline number.

Position

The other credibility-shrunk, when-mentioned signal.

Sentiment monitoring

Read the actual answers behind your sentiment score.