# Hyperize

> Hyperize builds the Agent Surface: content, structured pages, matching JSON data and functions AI agents can use. From finding answers to requesting quotes, booking and ordering.

Scoped index: Insights / Editorial. Site index: https://www.hyperize.ai/llms.txt, full index: https://www.hyperize.ai/llms-full.txt

## Insights / Editorial

- [Insights Hub — EN](https://www.hyperize.ai/en/insights): Research and field tests on AI visibility and usability: what AI agents can find and complete on company websites. With DAX 40 measurements and methodology.
- [Insights Hub — JSON](https://www.hyperize.ai/en/insights.json): Articles, existing answers, scope and next steps. Agent-ingestion endpoint.
- [Kann ein KI-Agent bei einem DAX-Konzern einkaufen? · DE](https://www.hyperize.ai/de/insights/articles/kann-ein-ki-agent-bei-einem-dax-konzern-einkaufen): Deutsche Datenstory zur Welle Q2 2026 Pilot des DAX 40 Agent Success Index. Nur 4 von 36 Konzernen bestehen den Test mit allen vier Agententypen (Deutsche Telekom, Qiagen, DHL, Infineon), bei adidas und Zalando kommt keiner weiter; am Ende des Rankings adidas, Allianz, Bayer, SAP, BMW; KI-Sichtbarkeit und Bedienbarkeit haengen praktisch nicht zusammen (r = -0,02); bei 14 Konzernen laeuft der Kauf laut KI-Antworten ueber Dritte. Alle 36 Zahlen als Tabelle, CSV und JSON, CC BY 4.0.
- [Artikel-JSON · DAX-Zahlen-Story DE](https://www.hyperize.ai/de/insights/articles/kann-ein-ki-agent-bei-einem-dax-konzern-einkaufen.json): Dichter Agent-Ingestion-Endpunkt: vier Kernsaetze (keyStatements), das volle 36-Zeilen-Ranking mit beiden Achsen, Ergebnis je Agentenart und Zaehlwerte, Vermittler-Anteile, Talent-Zaehlung, Zitierblock, Quellen.
- [Can an AI agent buy from a DAX company? · EN](https://www.hyperize.ai/en/insights/articles/can-an-ai-agent-buy-from-a-dax-company): English twin of the German data story on wave Q2 2026 Pilot of the DAX 40 Agent Success Index. Only 4 of 36 companies pass the test with all four agent types (Deutsche Telekom, Qiagen, DHL, Infineon), at adidas and Zalando none gets any further; the ranking ends with adidas, Allianz, Bayer, SAP, BMW; AI visibility and usability for agents are practically unrelated (r = -0.02); at 14 companies the purchase goes through third parties according to AI answers. All 36 numbers as table, CSV and JSON, CC BY 4.0.
- [Article JSON · DAX numbers story EN](https://www.hyperize.ai/en/insights/articles/can-an-ai-agent-buy-from-a-dax-company.json): Dense agent-ingestion endpoint: eight key statements, the full 36-row ranking with both axes, result per agent type and counts, intermediary shares, Talent counts, citation block, sources; alternateLanguage points at the German master.
- [DAX 40 Agent Success 2026: the numbers · EN](https://www.hyperize.ai/en/insights/articles/dax-40-agent-success-2026-the-numbers): Numbers article for Wave Q2 2026 Pilot of the DAX 40 Agent Success Index. All 36 brands scored on both axes in one ranked table (rank, ticker, sector, AI Visibility, AI Usability, score 0 to 10), Deutsche Telekom 7.0 to Brenntag 2.5, the three recurring patterns (Qiagen: ready but rarely found; Allianz: portals capture the category first; adidas: blocked at the door), the Talent lane kept separate, and a how-to-cite block. Dataset CC BY 4.0.
- [Article JSON · DAX 40 numbers 2026](https://www.hyperize.ai/en/insights/articles/dax-40-agent-success-2026-the-numbers.json): Dense agent-ingestion endpoint. The full 36-row ranking with both axes and evidence per brand, coverage counts, the three patterns, Talent-lane counts, citation record, sources.
- [Hyperize receives the BSFZ seal for in-house R&D. — EN](https://www.hyperize.ai/en/insights/articles/hyperize-receives-bsfz-seal): Company-News article. The BSFZ certified the R&D project behind the Hyperize platform under Germany's Research Allowance Act (issued to MING Labs GmbH, 2026-08-10; red variant = R&D at least partly in-house). Covers the BSFZ's three assessment criteria, the decision-vs-seal distinction, and the structured-data representation of the credential on the About page.
- [Article JSON — BSFZ seal](https://www.hyperize.ai/en/insights/articles/hyperize-receives-bsfz-seal.json): Dense agent-ingestion endpoint. Credential record (issuer, holder, certified project), key facts, sources.
- [We asked Grok Bot to find an office in Munich · EN](https://www.hyperize.ai/en/insights/articles/we-asked-grok-bot-to-find-an-office-in-munich): Field-test article. A consumer agent (Grok Bot, SpaceXAI) given a realistic Munich office-search brief: ImmoScout24 answers 401 behind a robots.txt that welcomes AI crawlers, Kleinanzeigen bans the agent's IP range before it can read the policy, Immowelt opens but serves counts and areas the agent cannot trust. Three refusal layers (published policy, network edge, content), all probes independently reproduced from datacenter and residential egress.
- [Article JSON · Grok Bot office field test](https://www.hyperize.ai/en/insights/articles/we-asked-grok-bot-to-find-an-office-in-munich.json): Dense agent-ingestion endpoint. Machine-readable probe table (per-host policy / edge / content), test parameters, the three-layer refusal pattern, what portals would need to serve delegated agents, sources.
- [KI findet Ihr Angebot. Kommt sie auch weiter?](https://www.hyperize.ai/de/insights/articles/the-agent-revenue-leak): Vier historische Einzeltests zeigen Hürden beim Website-Zugang, beim Lesen von Produktdaten und beim nächsten erlaubten Schritt. Sie belegen keinen verlorenen Umsatz.
- [Agent Revenue Leak · JSON DE](https://www.hyperize.ai/de/insights/articles/the-agent-revenue-leak.json): Article, sources and limitations.
- [Agent Revenue Leak · Markdown DE](https://www.hyperize.ai/de/insights/articles/the-agent-revenue-leak.md): Readable article with sources.
- [AI finds your offer. Can it take the next step?](https://www.hyperize.ai/en/insights/articles/the-agent-revenue-leak): Four historical individual tests show barriers to website access, product data and the next permitted step. They do not establish lost revenue.
- [Agent Revenue Leak · JSON EN](https://www.hyperize.ai/en/insights/articles/the-agent-revenue-leak.json): Article, sources and limitations.
- [Agent Revenue Leak · Markdown EN](https://www.hyperize.ai/en/insights/articles/the-agent-revenue-leak.md): Readable article with sources.
- [Ein Button bringt Ihr Angebot in die KI-Prüfung.](https://www.hyperize.ai/de/insights/articles/we-built-a-button-ai-closed-the-deal): Der CWO-Button kopiert ein Kurzbriefing für die eigene KI. Ein ausdrücklich erfundenes Hotelbeispiel zeigt, wie sich Angebotsangaben und offene Fragen prüfen lassen. Das belegt weder eine Buchung noch einen Verkaufsabschluss.
- [CWO practice · JSON DE](https://www.hyperize.ai/de/insights/articles/we-built-a-button-ai-closed-the-deal.json): Article, illustrations, sources and limitations.
- [CWO practice · Markdown DE](https://www.hyperize.ai/de/insights/articles/we-built-a-button-ai-closed-the-deal.md): Readable article with sources.
- [A button puts your offer in front of the buyer’s AI.](https://www.hyperize.ai/en/insights/articles/we-built-a-button-ai-closed-the-deal): The CWO button copies a briefing for the customer's own AI. An explicitly fictional hotel example shows how to check offer details and unanswered questions. It proves neither a booking nor a sales close.
- [CWO practice · JSON EN](https://www.hyperize.ai/en/insights/articles/we-built-a-button-ai-closed-the-deal.json): Article, illustrations, sources and limitations.
- [CWO practice · Markdown EN](https://www.hyperize.ai/en/insights/articles/we-built-a-button-ai-closed-the-deal.md): Readable article with sources.
- [Is your brand ready for AI agents? — EN](https://www.hyperize.ai/en/insights/articles/brand-ready-for-ai-agents): Foundational/Manifesto-tier article. Eight convictions, 9 animated charts, 4 coined concepts (Fossil Layer Test, Thick Server, Don't Make Me Compute, Agent-to-Agent Interface), Sources Block with HBR + 5 transaction-rail entries, 5-Q FAQ.
- [Article JSON — brand-ready](https://www.hyperize.ai/en/insights/articles/brand-ready-for-ai-agents.json): Dense article surface for agent ingestion. 4 coinedConcepts with reserved Concept Page @ids, full Sources Block, cluster context (foundational-anchor tier).
- [Can AI agents use German car configurators? — EN](https://www.hyperize.ai/en/insights/articles/four-german-cars-tested-by-ai-agents): Wave 4 sector article. BMW, Mercedes-Benz, Porsche, and Volkswagen given one job: configure a new car to a clear price. Only one configurator was readable by every agent class that ran; the others each broke in a different place.
- [Article JSON — four German cars](https://www.hyperize.ai/en/insights/articles/four-german-cars-tested-by-ai-agents.json): Dense agent-ingestion endpoint. Four-brand configurator benchmark (both axes), ground-truth prices, per-agent-class read grid, architecture findings (SSR / Shadow DOM / catalogue drift), sources.
- [Can AI agents buy from German pharma and healthcare brands? — EN](https://www.hyperize.ai/en/insights/articles/five-german-life-sciences-brands-tested-by-ai-agents): Wave 5 sector article. Bayer, Merck, Fresenius, Fresenius Medical Care, and Qiagen given one job each: reach the meaningful next step on the brand's own surface. Only the B2B catalogues close for every agent class; the consumer brand everyone knows has no buy button for the product an agent would recommend.
- [Article JSON — five German life sciences](https://www.hyperize.ai/en/insights/articles/five-german-life-sciences-brands-tested-by-ai-agents.json): Dense agent-ingestion endpoint. Five-brand benchmark (both axes), per-brand close-states, per-agent-class read grid, WAF / SSR / missing-buy-button findings, sources.
- [Can AI agents buy from German chemical brands? — EN](https://www.hyperize.ai/en/insights/articles/four-german-chemicals-brands-tested-by-ai-agents): Wave 6 sector article. BASF, Brenntag, Heidelberg Materials, and Symrise tested on one buying task each. Every brand is found; where the buy was measured, it closes off the brand surface (Amazon, the Baumarkt aisle, a quote-form wall). Symrise visibility-only this wave.
- [Article JSON — four German chemicals](https://www.hyperize.ai/en/insights/articles/four-german-chemicals-brands-tested-by-ai-agents.json): Dense agent-ingestion endpoint. Four-brand interception read (named / spec-on-surface / where-it-closes), AI Visibility per brand, derived usability where measured, public product + standards + market facts, sources.
- [Can AI agents buy German consumer goods on the brand's own site? — EN](https://www.hyperize.ai/en/insights/articles/two-german-consumer-goods-brands-tested-by-ai-agents): Wave 7 sector article. Beiersdorf (Eucerin SPF 50+) and Henkel (Persil Color) tested on one locked claim each. Both brand pages survive the agent's verification test; the purchase closes at dm, Rossmann, Edeka, REWE, or Amazon. The brand owns the claim; the shelf owns the checkout.
- [Article JSON — two German consumer goods](https://www.hyperize.ai/en/insights/articles/two-german-consumer-goods-brands-tested-by-ai-agents.json): Dense agent-ingestion endpoint. Two-brand interception read (named / claim-on-surface / where-it-closes), both axes (usability from the partial fleet profile), the three consumer-retail pathologies (open-page-no-close / marketplace / hard-block), Adidas anchor, sources.
- [Can AI agents buy from German aerospace and defence brands? — EN](https://www.hyperize.ai/en/insights/articles/three-german-aerospace-defense-brands-tested-by-ai-agents): Wave 8 sector article. Airbus (H145), MTU Aero Engines (PW1100G-JM MRO), and Rheinmetall (Lynx KF41): nothing in this sector sells by checkout, every close is a direct-sales RFQ. All three primes surface reliably (AI Visibility 43.9 / 52.5 / 49.4); variant fidelity and the RFQ pathway land with the next fleet pass.
- [Article JSON — three German aerospace and defence](https://www.hyperize.ai/en/insights/articles/three-german-aerospace-defense-brands-tested-by-ai-agents.json): Dense agent-ingestion endpoint. Three-prime visibility-only read (AI Visibility per brand, usability + composite pending a later wave), locked variant + RFQ close per brand, four heavy-industrial surface modes, Siemens anchor, sources.
- [Ask AI for a reinsurer. Munich Re comes up. The broker doesn't. — EN](https://www.hyperize.ai/en/insights/articles/three-german-insurers-tested-by-ai-agents): Wave 9 sector article on which reinsurers AI shortlists for European Property Cat XL Treaty capacity. Munich Re and Hannover Rück under the question every Chief Underwriting Officer asks before the 1 January renewal. The brands come up; the four broker houses that actually place the treaty do not.
- [Article JSON — three German insurers](https://www.hyperize.ai/en/insights/articles/three-german-insurers-tested-by-ai-agents.json): Dense agent-ingestion endpoint. Three-brand interception read (named / argument-surfaced / route-to-close), AI Visibility per brand, broker houses (Aon, Guy Carpenter, Gallagher Re), three intermediation modes, sources.
- [Ask AI for industrial equipment. The integrator doesn't come up. — EN](https://www.hyperize.ai/en/insights/articles/three-german-industrial-catalogs-tested-by-ai-agents): Wave 10 sector article on which German industrial OEMs AI agents shortlist for big-ticket equipment. Siemens Energy, Siemens Healthineers, and GEA Group come up; the EPC contractors, hospital procurement departments, and process-engineering firms that close the project do not.
- [Article JSON — three German industrial catalogs](https://www.hyperize.ai/en/insights/articles/three-german-industrial-catalogs-tested-by-ai-agents.json): Dense agent-ingestion endpoint. Three-OEM interception read (named / argument-surfaced / route-to-close), both axes per brand (usability as structural interim until a later wave), EPC / hospital-procurement / engineering-firm layers, Siemens AG anchor, sources.
- [Tires, trucks, and enterprise software: the last three — EN](https://www.hyperize.ai/en/insights/articles/the-last-three-dax40-brands-tested-by-ai-agents): Wave 11 census-close article on which DAX 40 brands AI agents can buy from. Continental (tires), Daimler Truck (Actros), and SAP (S/4HANA) close the first full Commerce-lane pass. Three industries, one pattern: the brand comes up, the retailer, dealer, or system integrator owns the close. Found, but not yet buyable.
- [Article JSON — the last three DAX40 brands](https://www.hyperize.ai/en/insights/articles/the-last-three-dax40-brands-tested-by-ai-agents.json): Dense agent-ingestion endpoint. Three-brand census-close read (named / argument-surfaced / route-to-close), both axes per brand (usability as structural interim until a later wave), retailer / dealer / integrator layers, Daimler Truck + SAP disambiguation traps, census-close milestone, sources.
- [AI agents built our websites. One human made the decisions. — EN](https://www.hyperize.ai/en/insights/articles/ai-agents-built-our-websites): Making-of Leadership article. How hyperize.ai and minglabs.com shipped with no design team and no dev team: design sandbox (Lovable), mechanical port into server-rendered Astro HTML, 12 page types with 34 automated checks per page, one human holding decisions and veto. Receipts: 17 days first-commit-to-launch, Lighthouse mobile 95, logged-out ChatGPT citing the index as source 1 of 3 eight days after launch, and the published failure (a ported demo form that dropped every lead for four days).
- [Article JSON — AI agents built our websites](https://www.hyperize.ai/en/insights/articles/ai-agents-built-our-websites.json): Dense agent-ingestion endpoint. Build pipeline at station altitude (sandbox / port / gates / launch), operating model, verified receipts with declared query classes, the lead-form incident, sources.
- [We automated our content. ChatGPT decides if it worked. — EN](https://www.hyperize.ai/en/insights/articles/the-proof-flywheel): Leadership article coining the Proof Flywheel: a content system in which the measurement that schedules the work also produces the proof that the work works. Five stations on one loop, three sensors, two human gates, one human strategic input, judged by whether live AI engines cite the output. Public ledger opened 2026-06-09 with two wins and two losses on day one (logged-out ChatGPT cited the index as source 1 of 3; Perplexity called the lane empty). Wins become published evidence; losses become pre-explained build orders. Vocabulary-led wins declared as such; every entry re-probed on a fixed date because citations decay.
- [Article JSON — the Proof Flywheel](https://www.hyperize.ai/en/insights/articles/the-proof-flywheel.json): Dense agent-ingestion endpoint. Coined-term definition, the machine at station altitude (five stations + three sensors + two gates + one human input + nesting model), the public ledger (dated, verbatim, reproducible entries with query class), measurement frame, scope, sources.
- [Can AI agents buy from your brand? Ask them.](https://www.hyperize.ai/en/insights/articles/can-ai-agents-buy-from-your-brand-seven-prompts): Seven prompts give initial clues about whether AI agents can find your offer, read your pages, perform a customer task and cite your evidence. A self-test provides findings, not a reliable score.
- [KI-Agenten selbst testen: sieben Prompts.](https://www.hyperize.ai/de/insights/articles/can-ai-agents-buy-from-your-brand-seven-prompts): Sieben Prompts zeigen erste Hinweise darauf, ob KI-Agenten Ihr Angebot finden, Ihre Seiten lesen, eine Kundenaufgabe ausführen und Ihre Belege zitieren. Ein Selbsttest liefert Hinweise, keinen belastbaren Score.
- [Seven prompts JSON EN](https://www.hyperize.ai/en/insights/articles/can-ai-agents-buy-from-your-brand-seven-prompts.json), [Sieben Prompts JSON DE](https://www.hyperize.ai/de/insights/articles/can-ai-agents-buy-from-your-brand-seven-prompts.json): Questions, conditions, documented limitations and sources.