01
What crawlers can read
HTML, robots, structured data, public copy, policy pages, and link paths.
Evidence-backed AI shopping readiness
Can AI-assisted buyers trust your product pages?
A bounded public-data audit that turns crawlable product evidence into a decision brief, remediation tickets, and a retest path — without pretending to guarantee AI rankings.
Manual fallback: jan@mg5.org. Static page, no server-side intake, no tracking pixel, no raw evidence pack in first touch.
Sample · Synthetic merchantReadiness packet
62/100
fix sprint
Variant facts, shipping caveats, and citeable policy paths need remediation before monitoring.
Structured dataobserved
Policy handoffneeds map
AI prompt probingmethod-labeled
The signal gap
Human-friendly pages are not automatically agent-ready.
The audit separates what exists publicly from what an AI-assisted buyer can safely cite, compare, and hand off.
02
What buyer agents can answer
Questions about variants, use cases, shipping, returns, trust, and exact-product routing.
03
What the team can fix
Owner-assigned remediation tickets with acceptance checks and a before/after delta certificate.
Methodology
A document-grade workflow, not a hype dashboard.
Bounded public evidence capture
Collect only scoped public pages and browser-visible facts.
Schema and content extraction
Separate product facts, policy facts, dynamic facts, and unsupported claims.
Method-labeled AI visibility checks
Optional point-in-time probes, never represented as platform ranking guarantees.
Readiness score and evidence matrix
Readable enough for leadership; specific enough for ecommerce, SEO, and development teams.
Remediation pack and retest
Tickets, owner hints, guardrails, and an honest before/after delta path.
Public-safe samplesynthetic
A sample you can inspect before paying.
The public sample shows artifact quality without exposing raw transcripts, real customer data, named competitor scoring, or local operator paths.
Rubric snapshot
| Dimension | Signals checked | Result |
|---|---|---|
| Answerability | Facts, variants, use cases | partial |
| Citation readiness | Policies, trust, canonical URLs | needs map |
| Purchase handoff | Exact routes, caveats | blocked |
| Retest path | Acceptance checks | ready |
Early-access scope
Starter screen or value pack.
Pricing remains a hypothesis until scoped: starter proof packet €499–999; value pack €1.5k–3k; monitoring only after the first audit proves value.
- One bounded site/category first
- Public data unless explicitly approved
- Decision brief plus implementation tickets
- Retest checklist and delta template
Guardrails
What this deliberately does not claim.
- No live AI ranking guarantee
- No traffic, revenue, or conversion guarantee
- No merchant endorsement implied by samples
- No legal/compliance certification
- No raw ZIP sent or linked without recipient review
FAQ
Short answers for skeptical buyers.
Is this SEO?
Adjacent, but narrower. It checks whether AI-assisted buyers and answer engines can extract, cite, compare, and safely route product information.
Do you guarantee ChatGPT or Google AI visibility?
No. Live checks are method-labeled, point-in-time evidence, not deterministic rankings.
Can agencies white-label this?
Yes, after one or two scoped pilots prove delivery quality and support load.
What data do you need?
For the starter packet, public URLs are usually enough. Permissioned feeds or analytics require explicit scope and data-handling approval.