PAGE: /method · EN · 10.10.2026 NEW VISION LAB · CLARITY & TRUST Review transparently. Move forward with purpose. Our approach to AI Discoverability and Business Building: traceable review, clear responsibilities and deliberate approvals. Source analysis and the Discover, Build and Monitor packages belong to AI Discoverability. Who you are, what you offer, who it is for, where you operate and why it matters: these questions establish a clear information foundation. This illustration does not show an actual AI system response. Document version Updated: 10 October 2026 · Website extension 0.4. Provider: New Vision Lab Ltd. 1. Define the scope Company, website, report language and relevant topics establish the starting point. If needed, AI prepares suitable questions. NVL checks scope and assignment before starting. 2. Examine public sources Discover currently uses OpenAI for source-based web research. It focuses on the supplied domain, actual links and clearly attributable evidence. The analysis does not log into customer accounts, make purchases or submit forms on third-party sites. Inaccessible, unreadable and unexamined content are treated as limitations. 3. Classify statements Provider claim: a statement by the business, not independent confirmation. Supported – AI classification: a proposed source relationship, still subject to human review. Unclear: the basis is insufficient for a reliable conclusion. Not examined: the investigation was not performed for that point. 4. Review and deliver Before approving a report, NVL reviews company attribution, traceability of findings, source references, contradictions, open facts and the practicality of recommendations. An AI classification alone is not approval. Unverifiable information remains visibly unresolved. For Build, NVL also reviews templates, intended placement, implementation steps and acceptance criteria. Only an explicitly released delivery appears in the client Hub. Questions and revisions remain traceable; a new internal AI draft does not replace an existing client delivery. Human review does not independently certify every business claim. Legal questions, tax information and specific security claims require the relevant qualified confirmation. What data AI receives Research receives the company name, website, report language, scope and questions. Approved business answers may supplement follow-ups. Contact emails and internal notes are not automatically included. Translations contain the relevant report text. Privacy reports are stored separately and are not sent to the analysis AI. For Build, AI receives the selected measures, their baseline, existing templates and instructions or client feedback deliberately supplied for that work. General Hub conversations and files are not automatically sent to AI in full. Privacy information What this does not establish A source review is not a representative visibility test across all AI systems. It guarantees neither rankings nor commercial outcomes and is not a security, legal or compliance certification. Comparisons document specific observations. Changes in models, questions, languages and source access are identified as limitations. NVL pilot · a concrete report excerpt New Vision Lab internal pilot · shortened excerpt from approved Build delivery v7 dated 27 September 2026. The underlying Discover assessment described an earlier website version. This is not a current live review. Review question: Is the scope of the three packages sufficiently specific for a purchase decision? Baseline finding: The package presentation described benefits and broad outcomes. Operational boundaries, output formats and the distinction between delivering templates and implementing website changes were insufficiently clear. Source: https://www.newvisionlab.net/plans-pricing. Priority recommendation: State included services, exclusions and deliverables for each package and align them with purchase information. Build output: developed text templates with intended placement, implementation steps and acceptance criteria. Example: “Build delivers copy-ready content and instructions. Your team or agency implements the changes.” Acceptance criterion: Do the package card, detailed service description and checkout agree on scope, output format and responsibility for implementation? The result will be recorded after publication and review. Limitations: This excerpt illustrates our workflow using our own business. It does not establish improved AI rankings, revenue effects or legal approval. Access limitations in the baseline review were not treated as proof of website defects. Report implementation. Verify the result. Clients can report individual Build tasks as implemented in the Hub and add a comment or evidence. This is separate from an NVL follow-up review. NVL records whether a change is visibly implemented, partially implemented or not verifiable. Later comparisons record website versions, questions, language, sources and relevant review conditions. Differences between AI runs alone do not prove a website improvement. Documented progress rather than a blanket visibility claim A follow-up review compares the approved scope, client feedback and publicly accessible content checked again. A client statement guides the review; it is not independent evidence. Content that cannot be read remains an explicit limitation. Monitor compares documented baselines within the agreed scope, distinguishing new, changed, unchanged and unverifiable findings. A different access route or question can limit comparability. Our own website is a pilot. Before-and-after examples show documented content changes and our workflow; they do not establish a guaranteed frequency of AI mentions or revenue growth. Focused functional checks on request For a separate website and enquiry journey check, we agree specific steps such as mobile usability, a form test submission or evidence of an enquiry in the authorised destination system. Public observations are distinguished from functions actually tested. An unmeasured enquiry and an undelivered enquiry are different problems. Tracking, consent settings and actual message receipt are checked only within the agreed scope and with suitable access. A source report alone does not verify a working technical integration. Business Building: review, agree, coordinate Business Building starts with a general enquiry. NVL provides the personal confidentiality undertaking before confidential discussions. We then review requirements, feasibility, specialist needs and potential collaboration. Services, responsibilities, costs, reviews and handover are individually agreed for each venture. Specialists receive only information required for their agreed tasks, with confidential content subject to the stated permissions. The new Business Building Hub is being developed in stages. AI-assisted breakdown of uploaded ideas into tasks, partner access to individual cards and automatic work allocation are not already promised features. A Business Building enquiry does not automatically start the existing AI Discoverability analysis. Business BuildingConfidentiality RELATED PAGES Home https://www.newvisionlab.net/en Business Building https://www.newvisionlab.net/en/company-builder AI Discoverability https://www.newvisionlab.net/en/ki-auffindbarkeit AI plans & FAQ https://www.newvisionlab.net/en/plans-pricing Client Hub · AI services https://www.newvisionlab.net/en/nvl-hub Confidentiality https://www.newvisionlab.net/en/confidentiality Service terms https://www.newvisionlab.net/en/terms Privacy https://www.newvisionlab.net/en/privacy Provider details https://www.newvisionlab.net/en/imprint Privacy concerns https://www.newvisionlab.net/en/terms#privacy-report CONTACT info@newvisionlab.net