Three-Score Model.

“One opaque score is a black box. Three transparent scores are a diagnosis. The difference is the difference between panicking and acting.”

— The Patnick perspective

What is the three-score model?

Patnick's three-score model decomposes AI visibility into three independent 0-100 dimensions — Demand (coverage breadth × real search demand), Clarity (entity consistency × cross-LLM consensus), Saturation (competitive crowding). Each dimension is transparent, research-backed, and mapped to a specific type of fix. Unlike single-number visibility scores from other tools, you can always see which component moved and why.

Why not use one composite score?

Composite scores hide their weights behind a single number. When the composite drops 10 points, you don't know whether it's coverage, engagement, entity consistency, or competitive pressure. The three-score model exposes each movement independently so you can respond with the right lever: Clarity drop → work on entity identity and schema; Demand drop → expand topical coverage or build historical data; Saturation rise → pivot positioning or differentiate on a topical border. One score tells you there's a problem. Three scores tell you which problem.

How is the Demand score calculated?

Demand = 0.5 × presence_rate + 0.3 × share_of_voice + 0.2 × (Google Ads volume normalized + Google Trends momentum), clamped 0-100. Presence rate is the fraction of probed queries where your brand entity is mentioned. Share of voice is your entity mentions divided by total entity mentions (you + competitors). Volume and momentum are normalized from external data. All weights are published — you can reproduce the calculation from raw probe data.

How is the Clarity score calculated?

Clarity = 0.5 × cross_LLM_consensus + 0.3 × sentiment_consistency + 0.2 × citation_density. Consensus is the fraction of queries where multiple LLMs agree on mentioning your brand (the single strongest entity consistency signal). Sentiment consistency is the inverse of sentiment variance across probes — lower spread means the LLMs agree on how to describe you. Citation density is the fraction of mentioning probes that include a URL back to your domain. Clarity is the formula expression of 'is your entity firmly resolved'.

How is the Saturation score calculated?

Saturation = 0.6 × min(avg_competitors_per_query / 5, 1) + 0.4 × ads_competition_index (when Google Ads data is available). A category where LLMs mention 5+ competing entities per query hits 100% on the first term. The Google Ads competition index (0-1 from Keyword Planner) provides the second signal. Saturation describes how crowded your topical border is — high Saturation means even strong topical authority has to compete hard, so pivoting or differentiating is more productive than pouring more content into the same niche.

Is a lower Saturation score better?

Yes. Lower Saturation means your topical border is open — LLMs don't default to a long list of competitors when answering queries in your category, so a well-positioned brand can dominate. High Saturation (75+) means every query surfaces many entities, and the winning strategy is usually differentiation or finding an adjacent border rather than adding more content. This inverts the traditional keyword-difficulty framing, which celebrates low-difficulty keywords; in the topical authority model, low Saturation is the empty real estate you should be grabbing.