Skip to main content
Every brand in Surfais has one headline number: the AIS Score. It’s a 0-100 measure of how visible your brand is in the answers ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews give to the questions your customers are already asking.

​
Overview

The AIS Score blends four signals into one composite: how often you’re mentioned, where you rank when you are, how positively you’re framed, and what share of the conversation you own against everyone else being named. It’s computed fresh from your live scan data — the actual answers the five platforms gave to your tracked prompts — designed to answer the question you actually have in mind when you look at it: “how visible am I?”
This is a different question from “how well-set-up am I to become visible?” — that’s your Readiness Score, a separate number from your audit. The two are never merged.

​
The four signals and their weights

The weighting is deliberate and asymmetric. Visibility carries 60% of the score because being mentioned at all is most of the battle. A brand that’s never named can’t be well-positioned, can’t be praised, and can’t hold a share of a conversation it isn’t part of — Position, Sentiment, and Share of Voice are all conditional on presence. Surfais calls this a presence-led model: the score rewards showing up first, and only then rewards showing up well.
Position and Sentiment can only be measured on the runs where your brand actually appears. To stop a handful of lucky mentions from banking a high score, both are credibility-shrunk toward a neutral midpoint until enough evidence has accumulated — see Why scores move for how this plays out for a brand that’s just getting started.

​
A worked example

Say your brand ran 20 scans this period and was mentioned in 10 of them — 50% presence — with a platform consistency of 40% (your mention rate averaged across all five platforms individually). When you are mentioned, your average rank among every brand named in the answer is 2nd, and the tone is generally positive (a raw sentiment average of 70). Across those same runs, competitors and other brands the AIs named collectively picked up 30 mentions to your 10.
Notice that Position and Sentiment don’t simply plug in their raw values (85 and 70). Both pass through the credibility shrink first — at 10 mentions they’re pulled some way toward the neutral midpoint of 50, which is why 85 becomes a credited 69.4. The breakdown drawer always shows you both numbers side by side.

​
What the AIS Score does not measure

The AIS Score tells you what AI platforms are saying about your brand today. It does not tell you why, and it does not tell you what to fix — that’s the job of the separate Readiness Score from your audit’s six-category assessment. A brand can have a strong AIS Score built on fragile foundations, or a weak AIS Score despite having done everything right structurally — the two are designed to diverge and mean different things when they do. Nor is the AIS Score a ranking against a fixed leaderboard — it’s your own brand’s measured visibility. Use Competitor Share of Voice to see how you stack up against specific competitors.

​
Score bands

Excellent — 80 to 100

Consistently surfaced across AI answers. You’re the default recommendation for the questions you track.

Good — 60 to 79

Frequently mentioned across AI platforms. Solid ground, with headroom on consistency or position.

Fair — 40 to 59

Visible for some prompts, with clear room to grow. Typically a presence problem more than a quality one.

Poor — 20 to 39

Largely invisible — most prompts don’t surface you yet. Start with Visibility.
Below 20 is Critical: effectively absent from AI answers. If you see “Awaiting first scan” instead of a number, that’s a different state entirely — no scan has run yet, so there’s nothing to score. A true 0 only appears after a scan finds no mentions anywhere.

Visibility

The 60%-weighted core signal: mention rate and platform consistency.

Why scores move

Bands, the credibility ramp for new brands, and common causes of movement.

AIS Score vs Readiness Score

Two different questions, two different numbers — never merged into one.

How scans work

Where the underlying data for every signal comes from.