PromptScore™: How We Measure Your Brand's Visibility Across LLMs

The Measurement Problem

In 2024, a CMO at a mid-sized European fintech asked their digital marketing team a simple question: "How does our brand appear when someone asks ChatGPT about neobanks?" The team had no answer. There was no metric, no dashboard, no industry standard. They ran a few manual tests, screenshotted the results, and wrote a two-page memo. A month later, the results had changed and the memo was obsolete.

This is the measurement problem that PromptScore™ was built to solve. Before you can optimize LLM visibility, you need to measure it — consistently, comparably, and at scale across every major AI platform.

PromptScore™ is a 0–100 visibility index, computed monthly, that tells you exactly where your brand stands in the AI search ecosystem.

The 5 Dimensions of PromptScore™

The score is not a single number drawn from a single data point. It is a weighted composite of five distinct dimensions, each targeting a different aspect of how your brand appears in LLM responses.

Dimension 1: Frequency (30% of score)

How often does your brand appear in LLM responses for the 50 most relevant queries in your sector? A brand that appears in 40 out of 50 target queries scores far higher on Frequency than one that appears in 8 of 50. This is the foundation — if you're not being mentioned, nothing else matters.

Example: For the query "best project management tool for remote teams", Notion appears in 44/50 LLM responses tested. Linear appears in 31/50. Basecamp appears in 9/50.

Dimension 2: Position (20% of score)

When your brand does appear, is it the first recommendation, the third, or buried in a qualifying paragraph? Position scoring rewards brands that are cited first — because user behavior in AI chat mirrors web behavior: first mentions get the most attention, the most clicks, and the most conversions.

Example: Booking.com appears first in 67% of travel booking queries. Expedia appears first in 18% of the same queries.

Dimension 3: Sentiment (20% of score)

Not all mentions are equal. A mention alongside "but their customer service has faced significant complaints" is worth less than a mention alongside "widely praised for its ease of use." Sentiment scoring uses NLP analysis across all detected mentions to compute a weighted positivity score.

Example: A fintech brand with high frequency but mixed sentiment (cited often, but frequently with caveats) will score well on Frequency and poorly on Sentiment — revealing a specific optimization lever: manage the narrative in third-party content.

Dimension 4: Multi-LLM Reach (15% of score)

A brand that appears strongly on ChatGPT but is absent from Perplexity, Gemini, and Copilot has a fragile visibility profile. Multi-LLM Reach scores the breadth of your presence — rewarding brands that appear across all major platforms rather than dominating one while missing others.

Example: A B2B SaaS that scores 82 on ChatGPT but only 31 on Gemini and 22 on Perplexity has a Multi-LLM Reach score of 45/100 — indicating that its content has been well-absorbed by OpenAI's training data but has not penetrated Google's or Perplexity's corpora.

Dimension 5: Sector Benchmark (15% of score)

Absolute scores are less meaningful than relative ones. A PromptScore™ of 55 in the luxury goods sector (where the average is 31) is exceptional. A 55 in the cloud computing sector (where the average is 72) is below average. Sector Benchmark normalizes your score against the distribution of your direct competitors.

How Your Score Evolves Over Time

PromptScore™ is computed monthly. Each month's score reflects the current state of LLM training data and retrieval pipelines — which update as models release new versions and incorporate more recent web content.

For most brands, meaningful score changes become visible within 60–90 days of starting a GEO optimization program. The fastest-moving dimension is typically Position (which can shift in 4–6 weeks as new optimized content gets indexed) while Frequency tends to move more slowly (8–12 weeks) because it requires the LLM's broader world model to absorb the new signals.

PromptAds provides a monthly report breaking down each dimension's change, identifying which optimization actions drove movement, and recommending the next 30 days of priority actions.

What the Numbers Mean

As a general benchmark, here is how to interpret absolute PromptScore™ values:

• 80–100: Top-tier visibility. Your brand is consistently cited first across major LLMs for your core queries. This is the position of category leaders.
• 60–79: Strong visibility with optimization opportunities. You appear regularly but not always first, and there are specific queries where competitors outperform you.
• 40–59: Moderate visibility. You appear in some queries but miss many. Third-party citation density is likely the primary bottleneck.
• 20–39: Weak visibility. You appear sporadically, often in secondary positions, frequently with neutral or qualified sentiment.
• 0–19: Effectively invisible. LLMs do not associate your brand with your sector at a meaningful level.

Getting Your Score

The first step is a free PromptScore™ audit, which delivers your baseline score across all five dimensions, your sector benchmark position, and a prioritized list of the top three optimization opportunities for your brand specifically. Most audits are delivered within 48 hours of the initial request.