{
  "format": "hyperize/v1",
  "@id": "https://www.hyperize.ai/en/methodology/agent-success-score.json",
  "type": "Methodology",
  "pageType": "Methodology",
  "name": "Agent Success Score — Hyperize Methodology",
  "description": "The Agent Success Score measures how well a brand lets an AI agent complete its human's task. One score per agent lane, from two measured gates: AI Visibility and AI Usability. The canonical Hyperize definition.",
  "url": "https://www.hyperize.ai/en/methodology/agent-success-score",
  "alternateLanguage": {
    "de": "https://www.hyperize.ai/de/methodology/agent-success-score"
  },
  "inLanguage": "en",
  "datePublished": "2026-07-02",
  "dateModified": "2026-07-11",
  "nextReview": "2026-10-11",
  "confidence": "A",
  "evidenceTier": "proprietary",
  "publisher": {
    "@id": "https://www.hyperize.ai/#organization",
    "name": "Hyperize",
    "url": "https://www.hyperize.ai",
    "parentOrganization": "MING Labs"
  },
  "primaryConcept": {
    "@id": "https://www.hyperize.ai/en/methodology/agent-success-score#defined-term-agent-success-score",
    "type": "DefinedTerm",
    "name": "Agent Success Score",
    "definition": "The Agent Success Score measures how well a brand lets an AI agent complete its human's task. One score per agent lane, computed from two measured gates: AI Visibility (does the agent find and recommend the brand) and AI Usability (once the agent arrives, does the surface let it act). Scored by real agents against frozen tasks, not by a checklist.",
    "extendedDefinition": "The subject is the brand, the agent is the visitor: the score records how much success-chance the brand gives each incoming agent. It measures the current setup as-is, so a low score is the optimization case, not an insult. Both gates must open for a task to succeed: a brand can be recommended and still un-buyable, or perfectly buyable and never found. Evidence is a confidence layer on the two gates, not a third sales axis.",
    "inDefinedTermSet": "https://www.hyperize.ai/en/methodology#defined-term-set"
  },
  "differentiator": "Readiness rates potential. Success counts outcomes.",
  "differentiationClaims": [
    "The Agent Success Score measures an outcome (did agents succeed), not a potential (is the surface ready).",
    "Real agents run frozen, pre-published tasks against the live surface; the score is not a file-scan checklist.",
    "AI Usability is derived from the observed per-agent-class access profile, never hand-rated; the profile is the truth, the number is a reproducible summary of it.",
    "The composite may only ever appear beside its decomposition (both gates plus the access profile); a composite alone is theater.",
    "Scores are per agent lane and never blended across lanes.",
    "The Agent Success Score is a brand-side metric, not the contact-center agent success rate: it measures how well a brand's own surfaces let an incoming AI agent finish a task, not how often support agents resolve a customer's issue."
  ],
  "disambiguation": {
    "notToBeConfusedWith": "contact-center agent success rate (a customer-support KPI)",
    "statement": "The Agent Success Score is a brand-side metric: how well a company's own surfaces let an incoming AI agent finish a task. It is not the contact-center agent success rate, which counts how often service agents resolve a customer's issue. Same words, different subject, different agent."
  },
  "model": {
    "gates": {
      "gate1": {
        "name": "AI Visibility",
        "question": "Does an agent find and recommend the brand for the task?",
        "method": "Query audit across the AI platforms buyers use, on unbranded task questions.",
        "weight": 0.2
      },
      "gate2": {
        "name": "AI Usability",
        "question": "Once the agent arrives, does the surface let it act?",
        "method": "An agent fleet runs the task on the live surface, from a plain reader to an autonomous operator. The score is derived from the per-agent-class access profile and close-state depth (usability-derivation/v1).",
        "weight": 0.7
      },
      "evidence": {
        "name": "Evidence",
        "question": "Is the result provable and consistent across methods?",
        "method": "Cross-method ground-truth consistency. A confidence layer, not a sales axis.",
        "weight": 0.1
      }
    },
    "composition": "Agent Success Score = (AI Visibility × 0.20) + (AI Usability × 0.70) + (Evidence × 0.10)",
    "compositionNote": "The composition is public; the scoring of individual runs into each gate is proprietary. The composite headlines a brand page only beside both axis bars and the access profile.",
    "convergence": "The two gates are measured by two independent instruments that share no machinery: a query audit (Gate 1) and an agent fleet (Gate 2). When both point at the same gap, the finding does not depend on either instrument being right alone.",
    "frozenTasks": "Every task is frozen and published before a wave runs, so the goalposts cannot move between measurements.",
    "liveExample": {
      "brand": "DHL Group",
      "aiVisibility": 46,
      "aiUsabilityDerived": 68,
      "evidence": 85,
      "compositeScale10": 6.5,
      "source": "https://www.hyperize.ai/en/dax40-index/brands/dhl"
    }
  },
  "lanes": {
    "rule": "One Agent Success Score per agent lane. Lanes are never blended into one cross-lane number.",
    "rationale": "A brand meets six kinds of agent, each on a different surface with a different job. A shopping agent and a press agent need different things; averaging them hides the signal.",
    "set": [
      {
        "id": "commerce",
        "label": "Commerce",
        "agentJob": "buy"
      },
      {
        "id": "talent",
        "label": "Talent",
        "agentJob": "apply"
      },
      {
        "id": "aftersales",
        "label": "After-sales",
        "agentJob": "resolve"
      },
      {
        "id": "procurement",
        "label": "Procurement",
        "agentJob": "source / quote"
      },
      {
        "id": "investor_relations",
        "label": "Investor",
        "agentJob": "research"
      },
      {
        "id": "press",
        "label": "Press",
        "agentJob": "research"
      }
    ],
    "coverage": "Every brand record in the DAX 40 Index shows a coverage meter for how many of the six lanes are measured. Commerce is measured first across all qualifying brands before any second lane opens."
  },
  "relatedConcepts": [
    {
      "name": "Task Selection Doctrine",
      "status": "published",
      "@id": "https://www.hyperize.ai/en/methodology/task-selection#defined-term-task-selection",
      "shortDescription": "The frozen, published tasks the score is measured against, and the rules that keep them fair across brands.",
      "url": "https://www.hyperize.ai/en/methodology/task-selection"
    },
    {
      "name": "Agent Surface",
      "status": "published",
      "@id": "https://www.hyperize.ai/en/methodology/agent-surface#defined-term-agent-surface",
      "shortDescription": "The machine-readable layer agents retrieve, cite, and act on. The asset a brand builds when the score shows the gap.",
      "url": "https://www.hyperize.ai/en/methodology/agent-surface"
    },
    {
      "name": "Agent Revenue Leak",
      "status": "published",
      "@id": "https://www.hyperize.ai/en/methodology/agent-revenue-leak#defined-term-agent-revenue-leak",
      "shortDescription": "The commercial diagnosis behind a low score: revenue lost to agents that find the brand but never reach the close.",
      "url": "https://www.hyperize.ai/en/methodology/agent-revenue-leak"
    },
    {
      "name": "DAX 40 Agent Success Index",
      "status": "live",
      "shortDescription": "The public dataset. Germany's largest brands scored with this metric wave by wave, each score beside its decomposition.",
      "url": "https://www.hyperize.ai/en/dax40-index"
    }
  ],
  "citationExcerpts": [
    "The Agent Success Score measures how well a brand lets an AI agent complete its human's task.",
    "Readiness rates potential. Success counts outcomes.",
    "Two gates, both must open: AI Visibility (find and recommend) and AI Usability (arrive and act).",
    "Agent Success Score = (AI Visibility × 0.20) + (AI Usability × 0.70) + (Evidence × 0.10)",
    "The profile is the truth; the score is a reproducible summary of it, never a hand rating.",
    "One score per lane. Never blended.",
    "The composition, not the derivation.",
    "The Agent Success Score is a brand-side metric, not the contact-center agent success rate."
  ],
  "faq": [
    {
      "question": "What is the Agent Success Score?",
      "answer": "The Agent Success Score measures how well a brand lets an AI agent complete its human's task. It is one score per agent lane, built from two measured gates, AI Visibility and AI Usability, and scored by real agents running frozen tasks against the live surface, not by a checklist."
    },
    {
      "question": "Is the Agent Success Score the same as a contact-center agent success rate?",
      "answer": "No. The Agent Success Score is a brand-side metric: how well a company's own surfaces let an incoming AI agent finish a task. A contact-center agent success rate is a support metric: how often service agents resolve a customer's issue. Same words, different subject and different agent."
    },
    {
      "question": "How is the Agent Success Score calculated?",
      "answer": "It is a weighted composite of two measured gates and a confidence layer: AI Visibility (0.20), AI Usability (0.70), and Evidence (0.10), scored per agent lane on a 0 to 100 scale and displayed 0 to 10. AI Usability is derived from the observed per-agent-class access profile, never hand-rated. How individual runs are scored into each gate stays inside the engagement."
    }
  ],
  "scope": {
    "publishes": [
      "The canonical definition of the Agent Success Score",
      "The two-gate model (AI Visibility + AI Usability) and Evidence as a confidence layer",
      "The composition weights (0.20 / 0.70 / 0.10) and the derived-not-hand-rated AI Usability rule",
      "The six-lane doctrine (per-lane scores, never blended, coverage meter)",
      "The frozen-task and two-instrument convergence principles",
      "Cross-references to sibling Methodology concepts and the public DAX 40 Index"
    ],
    "doesNotPublish": [
      "How individual runs are scored into each gate (the derivation)",
      "Query classification and trigger logic",
      "Task grids, prompt phrasings, and agent fleet configuration",
      "Run logs and raw retrieval transcripts"
    ],
    "rationale": "The metric belongs in the open. The derivation stays inside the engagement."
  },
  "engagements": [
    {
      "name": "Methodology Hub",
      "type": "upstream",
      "url": "https://www.hyperize.ai/en/methodology",
      "description": "The Hyperize glossary. Where every published concept term lives, indexed for AI assistants."
    },
    {
      "name": "DAX 40 Agent Success Index",
      "type": "applied",
      "url": "https://www.hyperize.ai/en/dax40-index",
      "description": "The methodology applied in public. DAX 40 brands measured wave by wave on the same doctrine."
    },
    {
      "name": "Founding Program",
      "type": "sprint",
      "cost": "EUR 4,500",
      "duration": "7 days",
      "url": "https://www.hyperize.ai/en/founding-program",
      "description": "Seven-day Agent Success Sprint. Where the score shows the gap, and what raises it."
    }
  ],
  "hasPart": [
    {
      "@id": "https://www.hyperize.ai/en/methodology/agent-success-score#defined-term-agent-success-score",
      "name": "Agent Success Score",
      "type": "DefinedTerm"
    }
  ],
  "sources": [
    {
      "id": "S1",
      "publisher": "Hyperize Internal — Methodology",
      "title": "Agent Success Score methodology (ars-methodology/v1.1) + usability-derivation/v1",
      "date": "May 2026",
      "path": "Hyperize HQ/knowledge/strategy.md",
      "type": "internal",
      "supports": "The two-gate model, the derived (never hand-rated) AI Usability rule, and the published composition weights."
    },
    {
      "id": "S2",
      "publisher": "Hyperize — Methodology",
      "title": "Task Selection Doctrine",
      "date": "May 2026",
      "path": "https://www.hyperize.ai/en/methodology/task-selection",
      "type": "internal",
      "supports": "Why tasks are frozen and published before a wave runs, so the goalposts cannot move between measurements."
    },
    {
      "id": "S3",
      "publisher": "Hyperize — DAX 40 Agent Success Index",
      "title": "DHL Group BrandScore (Commerce lane)",
      "date": "Q2 2026",
      "path": "https://www.hyperize.ai/en/dax40-index/brands/dhl",
      "type": "measured",
      "supports": "A live example of the score beside its decomposition: AI Visibility 46.0, derived AI Usability 68, Evidence 85, composite 6.5 of 10."
    },
    {
      "id": "S4",
      "publisher": "Hyperize — DAX 40 Agent Success Index",
      "title": "DAX 40 Agent Success Index (living dataset)",
      "date": "Q2 2026",
      "path": "https://www.hyperize.ai/en/dax40-index",
      "type": "measured",
      "supports": "The lane scaffold and coverage meter on every brand record: scores are per lane, lanes are never blended."
    }
  ]
}