Access and technical foundations
Crawl access, indexing signals, page responses, sitemaps, mobile usability and performance on the pages included in your scope.
HOW WE ASSESS YOUR WEBSITE
ARC-100: our public audit methodology
An evidence-led assessment of how your website can be found, understood and represented in search and AI answers. ARC-100 combines technical checks, documented answer testing and human review to turn findings into a practical action plan.
Public methodology · Updated October 7, 2026
What the assessment examines
Content and answer readiness
Whether services, locations, customer questions and useful answers are clear, accessible and supported by reliable information.
Structured data and business identity
Relevant schema, consistent business facts, service and location information, and connections between your website and public profiles.
Trust and evidence
Authorship, qualifications, contact details, policies and verifiable proof of work. We distinguish supported claims from statements that need evidence.
Observed AI visibility
Mentions, recommendations, citations, accuracy and competitor appearances in an agreed set of prompts and platforms. The report states exactly what was tested.
Risk, governance and measurement
Relevant public-facing privacy and accuracy signals, agreed AI-use controls and a practical plan for measuring changes. This is not a penetration test or legal compliance certification.
From scope to a usable roadmap
Agree the scope
Confirm your business goals, services, locations, pages, competitors, platforms and access permissions before testing.
Collect and validate evidence
Combine crawl and page diagnostics with manual inspection. Record URLs, observations, dates and supporting captures; check findings in their business context.
Test relevant customer questions
Use agreed discovery and comparison prompts. Record the exact prompt, platform, available model or mode, date, location assumptions, answer and cited sources. Separate branded questions from unbranded discovery.
Review and explain the score
Apply the agreed ARC-100 rubric, check conflicting signals and identify anything that could not be tested. Explain the evidence and limitations behind the result.
Prioritize and hand over
Provide recommended actions, expected benefit, estimated effort and dependencies. Agree who will implement changes and how a later check will compare results.
What does a score out of 100 mean?
The paid ARC-100 score summarizes findings within the agreed audit scope and methodology version. It is a diagnostic measure, not a probability that ChatGPT will recommend your business, a search ranking or a revenue forecast. Your report explains the applicable categories and the evidence supporting the result.
- Not tested is reported as not tested; it is not proof of a failure. Missing or inaccessible evidence is identified explicitly.
- The free automated score uses a separate, narrower method. It cannot be directly compared with the paid ARC-100 score.
- AI answers can change across prompts, dates, locations and sessions. An absence in our tested answers does not prove your business is never recommended.
- We do not lower scores to encourage a purchase. Findings and priorities follow the evidence.
What a documented finding looks like
ILLUSTRATIVE EXAMPLE — NOT A CLIENT RESULT
- Observation
- No Organization JSON-LD block was found in the fetched homepage HTML.
- Evidence
- Homepage URL, retrieval date, inspected HTML and validation output would accompany this finding in the client report.
- Interpretation
- This is a structured-data gap to investigate. It does not establish that the business is absent from ChatGPT answers.
- Action and verification
- Review existing markup, add appropriate accurate schema if needed, validate it and recheck the published page.
What you receive
- An executive summary and explained score
- An evidence register with tested pages and AI prompts
- A competitor comparison within the agreed test scope
- A prioritized remediation roadmap and verification plan
Scope, report depth, platforms and deliverables are confirmed before work starts. Implementation and ongoing monitoring are separate services unless your agreement includes them.
Questions before you book
Is ARC-100 an independent certification?
No. ARC-100 is AI Readiness Canada’s assessment framework, not a certification issued or endorsed by Google, OpenAI or a regulator.
Does the free check test live ChatGPT recommendations?
The automated website score alone does not establish live recommendation visibility. Read the snapshot’s stated test scope; any observed AI answer evidence must identify the platform, prompt and date.
Will a high score guarantee AI recommendations?
No. A score describes the assessed signals. Search rankings, citations and recommendations are controlled by the platforms and cannot be guaranteed.
Can my existing agency implement the recommendations?
Yes. The roadmap can guide your existing team or agency. Implementation is scoped separately.
Do you publish client information?
Client-specific evidence and reports are shared within the agreed engagement. A public case study requires client permission. This page explains the approach without exposing proprietary checklists or weighting rules.