Share of model is the percentage of AI-generated responses — across a defined set of queries — that cite or reference your brand, measured against the total responses for those queries. It is the primary metric for AI visibility programs, analogous to share of voice in traditional media buying or share of search in organic search strategy.

If ChatGPT, Gemini, and Perplexity collectively produce 100 responses to the queries most relevant to your category, and your brand appears in 14 of them, your share of model for that query set is 14%. That number — and how it changes over time — is the primary KPI of an AI visibility program.

Why Share of Model Is the Right Metric

Traditional digital marketing metrics — impressions, rankings, clicks, traffic — measure visibility on surfaces where users decide whether to engage. AI-generated answers are different: the AI system decides what to include, and the user receives a synthesized response that often doesn't require a click to act on.

This means the commercial impact of AI visibility operates upstream of traffic. A buyer who asks an AI assistant "which vendors should I consider for enterprise analytics?" and receives a response naming two or three brands has already formed a consideration set — before visiting any website, before running any search, before having any contact with a sales team.

Share of model measures your brand's presence in that consideration-forming layer. It is the only metric that directly captures AI visibility performance at the point where buyer intent is shaped.

How Share of Model Is Measured

Measuring share of model requires a systematic, reproducible sampling methodology across AI platforms. The core process involves four steps.

Step 1 — Define the query set. Select the 20 to 50 queries most relevant to your category: category-level queries ("best enterprise CRM"), comparison queries ("Salesforce vs HubSpot"), problem-statement queries ("how to improve sales pipeline visibility"), and branded queries ("what does [your brand] do"). The query set must be stable across measurement periods to make comparisons valid.

Step 2 — Sample AI responses. Run each query across ChatGPT, Gemini, Perplexity, Claude, and Copilot, recording which brands are cited in each response. Sample multiple times per query per platform (AI responses vary between sessions), and record citation presence as a binary signal: cited or not cited.

Step 3 — Calculate scores. For each platform and query category, calculate the percentage of responses that cited your brand. Aggregate to an overall share-of-model score. Repeat for each competitor in your analysis set.

Step 4 — Track over time. Repeat the sampling on a regular cadence — monthly for active programs, quarterly for baseline tracking. Changes in share of model against a fixed query set and competitor set give you a direct measure of program impact.

Share of Model by Platform

Different AI platforms require different measurement approaches because they retrieve content differently.

ChatGPT draws primarily from training data in its base mode. Share of model on ChatGPT reflects historical content authority — what was present in the training corpus and how authoritative it was. It changes more slowly than real-time retrieval platforms.

Perplexity performs live web retrieval on every query, making its citation behavior more responsive to current content and indexability signals. Share of model on Perplexity can change meaningfully within weeks of content improvements — making it the most useful platform for tracking short-term program impact.

Gemini draws from Google's search index and Knowledge Graph. Its share-of-model signals are closely correlated with traditional SEO and structured data quality, making it the platform most responsive to schema and E-E-A-T improvements.

A complete share-of-model measurement covers all five major platforms and distinguishes platform-specific performance from aggregate performance.

Share of Model vs Share of Voice

Share of VoiceShare of Model
Measures brand presence in media and advertisingMeasures brand presence in AI-generated responses
Calculated from impressions and ad spendCalculated from AI response sampling
Purchased or earned through media activityEarned through content authority and structured signals
Tracked via media monitoring toolsTracked via systematic AI response sampling
Correlates with brand awarenessCorrelates with AI-stage purchase consideration

Reporting Share of Model to Leadership

Share of model is a metric that translates well to CFO and CEO reporting because it maps directly to competitive position. A CMO who can show that their brand's share of model increased from 8% to 23% over a 90-day program — while a named competitor dropped from 31% to 19% — is demonstrating a measurable competitive shift in AI-mediated discovery.

The most effective share-of-model reports for leadership include: the baseline score at program start, current score, competitor scores at both points, the query categories where gains were largest, and the platform breakdown showing which AI systems are driving the improvement.

Frequently Asked Questions

What is share of model?

Share of model is the percentage of AI-generated responses — across a defined query set — that cite or reference your brand. It is the primary KPI for AI visibility programs, measuring your brand's presence in AI-mediated discovery before buyers visit any website or run any traditional search.

How is share of model different from share of voice?

Share of voice measures brand presence in paid media and earned media coverage. Share of model measures brand presence in AI-generated responses. The former is purchased or earned through media activity; the latter is earned through content authority, structured data, and entity clarity — and it measures a different, earlier stage of the buyer journey.

How often should share of model be measured?

Monthly measurement is recommended for active AI visibility programs. Quarterly measurement is sufficient for baseline tracking. The query set and competitor set must remain consistent across measurement periods for the data to be comparable.

Can I measure share of model with existing tools?

Standard SEO tools (Semrush, Ahrefs, Google Search Console) do not measure share of model. Measuring it requires a custom AI response sampling methodology — querying target platforms with a defined query set, recording citation outcomes, and aggregating the data. Brainpan.AI establishes this baseline as part of the AI Visibility Audit.

Get your share-of-model baseline

Find out your brand's current citation share across ChatGPT, Gemini, Perplexity, Claude, and Copilot — benchmarked against your top competitors.

Request AI Visibility Audit
Kevin Walsh, Founder of Brainpan.AI

Written and reviewed by

Kevin Walsh

Kevin Walsh is the founder of Brainpan.AI, where he builds AI visibility infrastructure, GEO/AEO strategy, schema systems, and citation optimization programs for brands that need to be retrieved, cited, and trusted by AI answer engines.