Point11
How We Score the Web
EngineeringMar 16, 2026

How We Score the Web.

The methodology, benchmarks, and scoring behind the Site Score.

Adam Fish

Adam Fish

Co-Founder & CEO

Robert Bernton

Robert Bernton

Co-Founder & CRO

Introduction

We built the Site Scanner to help brands see how agents experience their sites. It runs 40+ checks grounded in open standards from Google, OpenAI, and Anthropic — testing everything from structured data to bot accessibility to page performance. The result is a composite score, a plain-language summary, and a prioritized fix list. Brands that improve their Site Score rank higher in agent search, traditional search, and advertising auctions.

Site Score

apple.com

Top 36% of technology sites

ChatGPT0
Gemini0
Claude0
Grok0

Scan any site free at point11.ai.

When agents recommend brands, they draw from their training data, they use search engines, and they read your content. Starting this year, they will shop and purchase on behalf of humans. Sites that adapted for mobile devices a decade ago won. Now, that same shift is happening with agents.

Industry Benchmarks

The Site Scanner is a home inspection report for brands. It is built on open standards from Google, OpenAI, and Anthropic, and emulates the way agents rank brands. The score itself is not the sole factor of a brand's visibility to LLMs, but it is an important first step that a brand can control. We scanned thousands of sites to build industry benchmarks and help brands understand the condition of their site relative to industry peers.

Industry Benchmarks
p10–p90p25–p7550Median

Last updated March 2026. Benchmarks are refreshed monthly.

Scoring Methodology

Each scan runs 40+ checks across site performance, discoverability, and accessibility. The score is a composite metric using open standards like Google Lighthouse[1], published crawler specifications from OpenAI[2], Google[3], and Anthropic[4], and Point11's own signals. The result is a single score that reflects how well a site is built for both humans and agents.

Scoring Methodology
Top BrandIndustry Median

The gap is an opportunity for brands.

Advertising Impact

Poor site performance is a direct tax on advertising spend. Google uses Quality Score[5] and Meta uses Quality Ranking[6] to factor site quality into ad auction rankings. Brands with low Quality Scores pay more per click. In our experience working with the largest global advertisers, often 50% or more[7]. Improving your Site Score is one of the most direct ways to lower that cost.

This year, we expect up to 50% of purchases to be influenced by agents, up from 20% last holiday season[8]. Agents recommend brands, compare products, and complete purchases. The sites they can read, navigate, and transact on will get the traffic. The brands that are ready will win.