--- title: "Methodology — Hyperize" description: "The canonical Hyperize glossary. Concepts published as their own pages, indexed for AI agents to cite without paraphrase. Agent Surface, Task Selection, Context Window Optimization, and more." canonical: https://www.hyperize.ai/en/methodology lang: en last-updated: 2026-05-26 --- # Hyperize Methodology. Methodology The canonical glossary. Each concept lives as its own page, indexed for AI agents to cite without paraphrase. Most providers measure. [Hyperize builds](/en/insights/articles/ai-agents-built-our-websites) — and publishes the vocabulary the building runs on. Why this exists ## Search indexed pages. Agents read brands. Search engines indexed pages for humans. AI agents retrieve, evaluate, and cite. They surface a handful of sources where search used to return ten blue links. The brands they cite are the brands built to be read by them. This page indexes the canonical Hyperize vocabulary — terms a brand needs to understand the shift, and to be cited within it. Each one carries its own definition, its own disclosure boundary, and its own JSON sibling for agent ingestion. In this library ## The doctrine, piece by piece. Concept pages carry citable definitions AI agents can quote without paraphrase. The [DAX 40 Index](/en/dax40-index) is where those definitions land in public data — the doctrine, applied. Insights is where they land in narrative — the doctrine, told. The library grows as Hyperize publishes more of the methodology in the open. [ 01 · Concept · The category itself ### Agent Surface. The machine-readable layer of a brand that AI agents retrieve, interpret, cite, and act on. Three functions, four artifact types, one coherent corpus. Read ](/en/methodology/agent-surface) [ 02 · Concept · The metric ### Agent Success Score. How well a brand lets an AI agent complete its human's task. One score per lane, from two measured gates: AI Visibility and AI Usability. Readiness rates potential; success counts outcomes. Read ](/en/methodology/agent-success-score) [ 03 · Doctrine · DAX 40 Index v3 ### Task Selection Doctrine. Six rules, five failure modes, and the Direct Surface / Third-Party Interception split that decides what's fair to measure. Applied to Allianz, Mercedes-Benz, DHL, Bayer. Read ](/en/methodology/task-selection) [ 04 · Concept · Push, not pull ### Context Window Optimization. The third era of brand optimization. SEO ranks for crawlers. GEO cites for retrievals. CWO recommends when a buyer pastes evidence-bound content into ChatGPT, Claude, Gemini, or Copilot. Read ](/en/methodology/cwo-context-window-optimization) [ 05 · Concept · The diagnosis ### Agent Revenue Leak. Revenue lost to AI agents that find a brand but never reach the close. Two leaks: the order routes to an intermediary, or the last step is unusable. Read ](/en/methodology/agent-revenue-leak) [ 06 · Concept · The upstream leak ### Third-Party Interception. The brand is found, but a portal, dealer, or contractor takes the transaction first. A routing problem, not a visibility one. Read ](/en/methodology/third-party-interception) [ 04 · Index · Where the doctrine is applied ### DAX 40 Agent Success Index. The public dataset where the methodology lands. DAX 40 brands scored on two axes, AI Visibility and AI Usability, under frozen public task grids. More brands added each wave. Open ](/en/dax40-index) [ 05 · Editorial · Where the doctrine is narrated ### Insights. The editorial layer — research, field notes, and original analysis on how AI agents evaluate brands. Each cluster starts with one measurement. Two foundational articles live: Context Window Optimization (CWO) and the eight convictions on brand readiness for AI agents. Open ](/en/insights) [ 06 · Answers · Where the doctrine is asked ### Answers. The buyer-question library. One question per page, direct answer in the first 75 words, evidence anchored to the methodology or the DAX 40 Index. GEO vs SEO, AI-readiness diagnostics, and more Hyperize-buyer questions shipping weekly. Browse ](/en/answers) On scope The object, not the operating system. The architecture behind individual artifacts, the rules that govern each type, the scoring formulas, the measurement protocols, and the templates that produce them are proprietary. The category belongs in the open. The operating system stays inside the engagement. ## Three ways to engage. Entry ### Snapshot Free Agent Success Snapshot 50 queries across 4 AI platforms. Your brand, measured. Delivered within 48 hours. No commitment. [Get a Snapshot](/en#snapshot) Recommended Sprint ### Founding Program €4,500 · 7 days · Full clarity Seven-day Agent Success Sprint. Where agents find you, where they can't, what to fix first. [Start the sprint](/en/founding-program) Context ### About Hyperize Who's behind this Where the methodology came from. The team. MING Labs heritage. The DAX 40 Index program. [Read about us](/en/about) --- Canonical: https://www.hyperize.ai/en/methodology Machine-readable index: https://www.hyperize.ai/llms.txt