What this measures

Visibility is not enough. Answers need evidence.

This scorecard checks content operations behind reliable AI answers: ownership, coverage, structure, freshness, machine access, citations, safety, escalation, feedback, and in-product delivery.

0 of 10 answered

01 · Source controlWhere does your approved support truth live?

Across inboxes, chats, and people’s headsIn docs, but ownership is unclearIn one reviewed, owned source of truth

02 · Question coverageHow closely does content match real customer questions?

We write what seems usefulWe use some tickets and search termsWe routinely turn support demand into content

03 · Answer structureCan each page answer one clear intent without extra context?

Most pages are long or cover mixed topicsSome pages are focused and scannablePages use direct answers, steps, and clear headings

04 · FreshnessHow do you keep answers aligned with product changes?

Updates happen after customers complainOwners review important pages occasionallyContent has owners, review dates, and update triggers

05 · Machine accessCan search engines and AI agents reliably access your help content?

Content is hidden, fragmented, or JavaScript-dependentPages are crawlable but lack consistent discoveryPages are crawlable, linked, indexed, and machine-readable

06 · EvidenceCan an AI answer point users back to supporting evidence?

Sources are not trackedLinks exist, but citations are inconsistentAnswers map to approved, identifiable sources

07 · BoundariesWhat happens when source material cannot support an answer?

AI is expected to answer anywayUsers receive a generic fallbackAI declines clearly and explains the next step

08 · Human handoffCan unresolved questions reach a person with context intact?

Users start again in another channelA contact route existsConversation, user, and product context transfer together

09 · Feedback loopDo you know which questions remain unanswered or unhelpful?

We do not measure thisWe inspect conversations manuallyWe track gaps, outcomes, and content improvements

10 · In-product deliveryCan users get answers where they become stuck?

They must leave the product or open a ticketWe link to a separate help centreSearch or grounded help is available inside the product

Free founder review

Turn your score into an action plan.

Share your site and result. FAQ Hub will review your public support journey and send three prioritised changes for stronger AI answers. No generic PDF.

Manual review of your public support content
Three high-impact readiness fixes
No obligation to buy FAQ Hub

Will AI agents trust your support content?

Visibility is not enough. Answers need evidence.

Turn your score into an action plan.