Share of Model: The Metric That Replaces Page Rank

Harvard Business Review introduced a concept that captures the new competitive landscape: Share of Model.

It measures how prominently your brand appears in LLM responses across three dimensions: 1. Mention frequency - how often you're referenced 2. Position prominence - where you appear in responses 3. Sentiment - how you're described in AI-accessible sources

This is the new page rank. And most brands don't even know their score.

How To Measure It

Run these queries across ChatGPT, Claude, and Perplexity. Document the results.

Query 1: Category leadership "What are the best [product category] for [common use case]?"

Are you mentioned? In what position? With what language?

Query 2: Direct brand knowledge "Tell me about [Your Brand] [Product Name]."

Does the AI have accurate information? Where do gaps exist?

Query 3: Comparison framing "Compare [Your Brand] vs [Competitor]."

What attributes does the AI use? Are they accurate? Do they favor you or the competitor?

Query 4: Purchase intent "I need to buy [product]. What should I get?"

This is the money query. If you don't surface here, you're losing sales.

What Drives Share of Model

Research suggests a weighted model:

- Citation frequency (~25%): How often you're referenced in sources agents consult - Position prominence (~20%): Where you appear in responses (first vs. mentioned later) - Domain authority (~20%): Are you cited by high-trust sources? - Content freshness (~15%): LLMs favor recent sources - Structured data quality (~20%): Can agents parse your product information?

There's a ~0.65 correlation between top-10 Google rankings and LLM mentions—the two disciplines aren't entirely separate. But correlation varies by platform. It's high for Google's AI Overviews but low for ChatGPT, which often generates responses without fetching online content at all.

The Visibility Gap

A brand with 10% market share can have 50% AI visibility if they've optimized. A brand with 50% market share can have 0% AI visibility if they haven't.

The data is stark. SimilarWeb found ChatGPT accounted for 16% of Zara's inbound traffic between June-August 2025. 8% of traffic to H&M and Aritzia. AI assistants became a primary discovery channel in under a year.

The Compounding Problem

AI visibility compounds. Agents learn from past recommendations. Brands visible today get more data about what converts, which improves future visibility.

If your competitor figured this out six months ago, they have a six-month head start on that flywheel.

What To Do This Week

1. Run the four queries above. Document where you stand. 2. Check your competitors with the same queries. Who's winning? 3. Audit your structured data. Is it complete? Consistent? 4. Look for content gaps. What questions does AI struggle to answer about you? 5. Set up monthly tracking. Share of Model should be a recurring metric.

Most brands haven't measured this. The window to become a first mover in your category is still open.

But check. Today.